THE COLORADO EXPERIENCE
WITH INSPECTION / MAINTENANCE
DATA HANDLING:
MACHINE READABLE FORMS
MAY, 1982
PREPARED FOR:
UNITED STATES y*9*8":
ENVIRONMENTAL £ "
PROTECTION
AGENCY
PREPARED BY:-
INSPECTION/MAINTENANCE STAFF
| 2565 PLYMOUTH ROAD
f ANN ARBOR, MICHIGAN 481O5
NATIONAL CENTER FOR VEHICLE EMISSIONS CONTROL AND SAFETY
DEPARTMENT OF INDUSTRIAL SCIENCES
COLORADO STATE UNIVERSITY
2O5 L.L. GIBBONS
FORT COLLINS, COLORADO 8O523 —
-------
-------
The Information 1n this document has been funded wholly or 1n part
by the United States Environmental Protection Agency under assistance
agreement number T901383-01 to Colorado State University, It has been
subjected to the agency's peer and administrative review and it has been
approved for publication. The contents reflect the views and policies
of the agency.
ACKNOWLEDGEMENTS
The National Center wishes to acknowledge the willingness of the
AIR Program Staff to provide Information and documentation and to
check the text for accuracy.
Colorado Department of Health:
Richard Barrett, Mobile Sources
Leo Carroll, Mobile Sources
Richard Fawcett, Data Services
Gerald Gallagher, Mobile Sources, Program Director
Kenneth Nelson, Mobile Sources
Nadine Quigley, Mobile Sources
Steve Sargent, Mobile Sources
Colorado Department of Revenue:
Brian Her, Program Administrator
The information provided by Colorado State University staff members,
especially Dr. Elmer Remmenga, 1s appreciated, as is that of the
National Computer Systems (NCS) representative, Tom Gallogly.
111
-------
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS iii
TABLE OF CONTENTS ........ v
LIST' OF TABLES ix
LIST OF FIGURES . . . 1x
1.0 INTRODUCTION 1
1.1 Data Handling ..... 1
1.2 Program Description 2
2.0 PRE-IMPLEMENTATION PLANNING 7
2.1 Sources of Planning Information 7
2.2 Identifying Data Needs 7
2.3 Sampling 11
2.4 Data Collecting and Handling Alternatives 11
2.5 Additional Investigations ... 14
3.0 IMPLEMENTATION OF DATA SYSTEM 21
3.1 Inspection Form 21
3.2 Mechanics' Training 32
3.3 Data Input 34
3.4 Readability 39
4.0 DATA ANALYSIS 43
4.1 Purpose and Functional Elements 43
4.2 "Blue System" Accuracy Checking ..... 43
4.3 "Green System" Accuracy Checking . 44
4.4 "Blue System" Information Producing . . 59
4.5 "Green System" Information Producing 59
-------
TABLE OF CONTENTS (continued)
Paqe
/
4.6 Staffing, Equipment, Costs, and Other Computer
Considerations 62
4.7 Administrative Detail 64
5.0 SUMMARY . 67
5.1 Description 67
5.2 Critical Aspects of the AIR Program 68
APPENDICES , 73
Appendix A I/M Program Sub-System, Arthur Young &
Company Report 73
Appendix B Statistical Analyses and Sampling Techniques
Documents 103
Appendix C Vehicle Emissions Control Study Guide - AIR
Program, Inspection Form Section 117
Appendix D Inspection Forms, July-December ("Blue") and
January-Current ("Green") 139
Appendix E NCS Forms Design Guide 161
Appendix F NCS Price List, January 1982 201
Appendix G "Blue System" Reports ... 209
Appendix H Overview of Operation of SPSS . . 223
REFERENCES CONSULTED 237
vii
-------
LIST OF TABLES
Page
Table 1: Estimations of Data Volumes 8
Table 2: Data Collecting and Handling Alternatives 12
Table 3: Fatal Error Identifications 46
Table 4: Questionable Error Identification . 48
LIST OF FIGURES
Figure 1: Forms Contrast 24
Figure 2: Data Input - January 1, 1982 to Present 35
Figure 3: Data Analysis - January 1, 1982 to Present 45
Figure 4: Processing System Error Messages . 47
Figure 5: Error/Performance Report (Mock-up) . 55
Figure 6: Batch Summaries (Pie Chart Graphics, Mock-up) .... 58
Figure 7: Data Analysis Projections 61
ix
-------
1.0 INTRODUCTION
1.1 Data Handling
This document will describe the inspection/maintenance (I/M) data
handling system for the Automobile Inspection and Readjustment (AIR)
Program operated by the State of Colorado. Machine readable forms
(also called optically read or scanned forms) play a key role in that
program's inspection data system. The intended purpose of this report
is to provide detailed descriptive information for the consideration
of other I/M programs. No data handling system should or can be
exactly duplicated in another program. However, understanding the
developmental process of one state's system does provide opportunities
for other programs to select and refine those components which suit
their unique needs.
Data handling is a crucial component of an inspection/maintenance
program. It provides a "window" into that enigmatic system of
stringency factors, types of repairs, calibrations, tailpipe inspec-
tions, reductions, costs, consumer protection, and quality control.
The accuracy and completeness of data and the design and number of the
analyses determine the validity and scope of the answer to the
question, "How well is the I/M program operating?"
Data handling systems are designed through a series of compro-
mises. Ideally, every element of applicable data would be collected,
and each would be 100% accurate. The analyses would require only
these data and, coping with all contingencies, they would accurately
answer every question for the many constituents of the I/M program.
If this ideal system were even achievable, the cost would be prohibi-
tive. At a more realistic level, compromises are struck:
-------
• completeness of data vs. costs of collection,
• completeness of data vs. accuracy of data,
• accuracy of data vs. costs to handle data,
• completeness and accuracy of data vs. ease of use,
• types of analyses vs. time requirements,
• types of analyses vs. costs of analyses, and
• types of analyses vs. value of information in analyses.
The result should be a data handling system which collects essential
information with a moderate amount of cost and effort, which has a
determinable measure of accuracy at or above an accepted standard, and
which analyzes the data accurately and usefully. Each data handling
system develops out of the needs of the I/M program it serves,
balancing the ideal against the practical. Describing the data
handling system developed by Colorado reports the efforts of one state
to design a means of measuring the efficiency and quality of its I/M
program.
1.2 Program Description
Senate Bill 52 (S.B. 52) established the Colorado Automobile
Inspection and Readjustment (AIR) program. The bill outlines an
annual, decenteralized inspection and adjustment program. Included in
the program are all 1968 and newer model passenger cars and light-duty
trucks in the following categories: 1968 through 1978 model years
with a gross vehicle weight rating (GVWR) of 6,000 pounds or less, and
1979 and newer model years with a GVWR of 8,500 pounds or less. All
diesel and propane powered vehicles, as well as motorcycles, are
exempt. The program area originally involved a nine county area along
the Colorado front range, from Colorado Springs in the south to Fort
Collins in the north. One county was later excluded, leaving the
-------
program area the entire counties of Boulder, Denver, Douglas and
Jefferson and urban areas of Adams, Arapahoe, El Paso, and Larimer
counties.
Vehicles passing the idle test are issued a window sticker marked
"pass." Five mandatory adjustments to manufacturers' specifications
are made on vehicles for which emissions exceed state cutpoints, then
a retest is conducted. Only on 1981 and newer vehicles are repairs
mandated ($100 maximum) and this not until July 1, 1982. Vehicles
which comply with the program requirements but are still not able to
pass the state cutpoints are issued a window sticker marked "adjust."
Inspections and adjustments are conducted at over 1,000 licensed pri-
vate garages by more than 3,000 licensed mechanics. Adjustment by the
vehicle owner is also allowed. The window sticker marked "adjust"
cannot be issued unless the adjustments are done by a licensed
emissions mechanic. The maximum failure rates allowed by statute (S.B.
52) are 40% for 1968-1974 model year vehicles and 30% for 1975 and
newer model year vehicles.
The AIR Program is regulated by the Air Quality Control Commission,
an interdisciplinary committee appointed by the legislature. Two
state agencies, the Department of Health (CDH) and the Department of
Revenue (CDR), share in the operational aspects of the program.
The Department of Health, Air Pollution Control Division, Mobile
Sources Section is the agency responsible for technical aspects of the
AIR Program. It operates three Vehicle Emissions Technical Centers at
which specialized tests can be conducted. It coordinates all data
handling; conducts public awareness activities; makes recommendations
-------
for emissions standards; handles the mechanic training and qualifica-
tion program; and maintains contact with the mechanics in all matters
related to 5.B. 52 and technical aspects of the AIR Program. The
Mobile Sources Section functions as a technical staff to the Air
Quality Control Commission. The Department of Revenue, Motor Vehicle
Division is concerned primarily with mechanic and station licensing
(issue, deny, cancel, suspend, revoke) and enforcement. Splitting
operational responsibilities between two agencies calls for special
data handling efforts to ensure that each agency, as well as the
Commission, receives the information it requires.
The development of Colorado's AIR program has had three distinct
phases to date. A limited voluntary inspection program (phase one) had
been operating since 1974, from which emissions data on about 20,000
vehicles had been gathered. The first mandatory inspections and
adjustments (phase two) began July 1, 1981, for government fleets, new
cars, vehicles changing ownership, and vehicles registering for the
first time in the program area (new residents). This phase was expected
to encompass about 80,000 vehicles over its six month span. In
actuality, closer to 140,000 vehicles were inspected. The inspection
form for the phase was printed with blue ink, thus providing a convenient
designation for the data collecting and handling: the "blue system."
With only 10% of the anticipated volume for the annual, expanded
program, this July-December period functioned as a "pilot test" for
the full-volume program. It allowed program staff opportunities to
test procedures, cutpoints, and data analysis programs. Implementation
of the full scale program began January 1, 1982, for an estimated 1.4
million vehicles in the program area.
-------
This constitutes phase three, also called the "green system" after the
green inspection form. The use of phases - the change of ownership and
fleet I/M program (phase two) and the full volume program (phase
three) - was a designed element intended to ease program implemen-
tation. The changes instituted for the "green system" frequently
resulted from problems and solutions identified during the "blue
system."
-------
2.0 PRE-IMPLEMENTATION PLANNING
2.1 Sources of Planning Information
Sources of information used as input for pre-implementation planning
included:
• data obtained in the seven-year old voluntary inspection
program,
• data accrued during several contracted studies completed for
the State of Colorado over approximately ten years,
• the data needs experienced by other I/M programs, and
• the data needs of this program as outlined in S.B. 52 and the
State Implementation Plan (SIP).
2.2 Identifying Data Needs
In 1980 the Colorado Department of Health (CDH) sought assistance
in evaluating the Air Division's entire data system. Arthur Young &
Company of Dallas, Texas, was contracted to undertake the study. The
project was completed in about eight months at an entire project cost
of approximately $130,000. Included in the study was a section on I/M
data needs, identifying the data requirements of the AIR Program.
This section was subcontracted to Radian Corporation because of
Radian's experience with I/M programs. Sections of the report dealing
with the AIR Program appear in Appendix A of this document. The
following raw data for the phases of the AIR Program emerged and are
outlined in Table 1. The "Estimated" column states rough estimates of
the total number of data elements which were expected to be processed
annually for the "green system." The "Projected from Actual" column
includes later figures calculated from actual inspection volumes. These
latter figures are sketchy. Knowing actual data element quantities has
not been essential to program management and the figures have not been
officially tallied.
-------
Table 1: ESTIMATIONS OF DATA VOLUMES
Type
Vehicle Identification
* License number
Vehicle Characteristics
• Model year
' Make
* No. of cy 1 Inders
Administrative
* Date of test
" Inspection station number
" Inspection station number
- for adjustment only
* Mechanic number
' Mechanic number - for adjustment only
' Certificate Issued
Cost
Inspection cost
Adjustment cost
Repair (labor) cost
Repair (parts) cost
Repair cost (total)
Inspection
Initial Test Visual Inspection (subcategory)
* Presence of catalytic converter
* Presence of fuel filler neck restrlctor
• Presence of AIR system
* Integrity of exhaust systems
Initial Test Emissions (subcategory)
* CO percentage
• CO pass/fal 1
' HC parts per mi 1 1 Ion
• HC pass/fall
Retest Visual (subcategory)
• Presence of catalytic converter
" Presence of fuel filler neck restrlctor
* Presence of AIR system
* Integrity of exhaust system
Retest Emissions (subcategory)
* CO percentage
' CO pass/fail
* HC parts per ml 1 1 Ion
• HC pass/fa! 1
DATA VC
Estimated3
for Green
1.5
1.5
1.5
1.5
1.5
0.1 (6.7%)
1.5
0.1
1.5
1.5
0.5
0.1
0.1
e
1.5
1 .5
.5
.5
.5
.5
.5
1.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
LUMES (In mil
Projected
Blueb
.14
.14
.14
.14
.14
.14
.024 (17*)
.14
.024
.14
.14
.024
d
d
e
.149
.14
.14
.14
.14
.14
.14
.14
9
g
9
9
.024
.024
.024
.024
Ions)
from Actual
Green
1.4
1.4
1.4
1.4
1.4
1.4
0.29 (21$)c
1.4
0.29
1.4
1.4
0.29
e
e
-------
Table 1: ESTIMATIONS OF DATA VOLUMES
Type
Repair
Voluntary Repair
Tune-up
Carburetor
Air cleaner
Choke
Other
Home adjust (subcategory )
CO percent
CO pass/fal 1
HC PPM
HC pass/fal 1
Overall CO and HC pass/fall
DATA VC
Estimated3
for Green
0.1
0.
0.
0.
0.
0.
0.
0.
0.
0.
LUMES (In mil
1-TojecTed
Blueb
1
I
I
[
i
I
I
e
I
e
1
Ions)
from Actual
Green
e
e
e
e
e
e
e
e
e
e
e
aArthur Young & Company, Colorado Air Pollution Control Division Information System
[draft for review and correction by sponsor), tn.p.: n.p., 7 August 1981), Section 2.5.8,
pp. 9-11.
'"'Blue system" was six months; "green system" Is annual beginning January 1982.
cActual stringency Is 17$ as of 4/1/82. The standards will become more stringent and
Increase the failure rate to roughly 25$. 21$ average for the year Is used here.
'•'Repairs labeled "required" on blue form but were actually only voluntary. Repair Is
for 1981 and newer models and will not be mandatory until July 1, 1982.
eData type not collected or estimated.
'This figure will be a small percentage of the 1981 and newer vehicle population as
the number of Inspected 1981 and newer vehicles Increases. These vehicles are those which
stlI I fat I after adjustment.
9A visual Inspection was conducted for all vehicles under the "blue system." Failure
data has not been analyzed.
hUnder the "green system" the visual Inspection is for 1982 and newer models. Because
new vehicles are automatically passed without an Inspection when new, 1982 models will not
be Inspected until 1983. First Inspection figures will be 0$ for 1982, growing to an
estimated 50$ of the volume of inspections by 1987. Economic factors could skew the
projections that 50$ of the vehicle population in 1987 will be 1982 or newer. A failure
rate is difficult to estimate and is a measure of the amount of tampering occurring.
'NO projections available.
-------
The discrepancies between the estimated volumes and the volumes
projected from actual inspection data are significant. This does not
deny the value of pre-implementation study, or even a study completed
by an outside contractor. It does suggest that figures generated
early in planning stages should be used only as indicators of
magnitude.
Identifying data needs for the AIR Program has been handled pri-
marily by CDH Air Pollution Control Division staff members. They are
familiar with the history of I/M in the state and thoroughly acquain-
ted with procedural and administrative aspects of this program and I/M
in general. The primary responsibility of one member, Kenneth Nelson,
is data handling. His background includes data processing as it per-
tains to air pollution, and this familiarity with data processing con-
cepts is beneficial when working with CDH Data Services staff members.
It is this latter section, headed hy Robert Little and under the imme-
diate supervision of Richard Fawcett, which handles actual computer
programming and systems design.
Identifying data needs is a constant process of reassessment. As
legislation and regulation altered the program design, the identified
data needs would require re-evaluation to be sure that type or volume
was not affected. At each juncture a decision also needed to be made
about the value of the information gathered and the effort required to
obtain it. Did one outweigh the other? Should a compromise be
struck? Identifying data needs continues, even when the program is
operational, albeit to lesser degrees. To be efficient, only essen-
tial data should be gathered, but a constant redefinition of
"essential" shows an awareness of the kaleidoscopic aspect of opera-
tional I/M programs.
10
-------
2.3 Sampling
The anticipated volume of inspection data prompted Colorado to
consider using data sampling techniques rather than the entire volume
of data. Sampling is expected to produce less precise results than
full volume analysis, but by greatly reducing the amount of data
handled, costs are also reduced. The risks of sampling include
possible bias in the collected data, thereby producing inaccurate
results, and insufficient cost savings to balance the expected
decrease in accuracy. Dr. Elmer Remmenga, professor of Statistics,
Colorado State Univesity, proposed a 25% sample size of a 1.4 million
records base. Even at that, the sample would still encompass about
350,000 inspection reports annually. In Dr. Remrnenga's October 9, 1980
report to the Department of Health he suggests:
With equal sample sizes, it would be very easy to obtain the
probability that each station is out of compliance each sampled
month for percent failed and change in HC and CO.
Optical scanner records leading to an overall 80-90 percent
sample would be preferable for efficiency, economics, and
statistics.^
2.4 Data Collecting and Handling Alternatives
In Dr. Remmenga1s estimation, a seni-automated data collection
process using the entire database and with a 80-90% accuracy level
was preferable to using sampling techniques. Additional research
pointed out that using a semi-automated system and all the data
records was not outrageously more expensive than a semi-automated
^Elmer Remmenga, unpublished report to Colorado Department of
Health, "Sample Size and Sample Plan Considerations for Evaluating
Emission Test Records for Consumer Protection," 9 October 1980, p. 2.
The report appears in its entirety in Appendix B of this document.
11
-------
system accessing only a sample of the collected data. Perhaps more
importantly, the accuracy of the analyses using all the data would be
greater than if only a sample of the data were used. Using machine
readable forms, hand-coded by the inspecting mechanic, appeared to be
a viable option in the progression of data collection and handling
alternatives (Table 2).
The following five alternatives were considered by the CDH staff:
Table 2: DATA COLLECTING AND HANDLING ALTERNATIVES
COLLECTING
HANDLING
inual
1.
2.
3.
4.
5.
homated
At Stations
handwritten
handwritten
hand coded
data tape in analyzer
and manually keyed
data
data tape in analyzer
and manual ly keyed
data
From Stations
by AIR staff
by AIR staff
by AIR staff
by AIR staff .
data transfer over
phone 1 ines
Submission
manual ly tabulated
keypunched
machine read
computer
computer
Ana lyses
manual
computer
computer
computer
computer
1. A totally manual system was out of the auestion considering
the data volume. The necessity for computer analysis was
obvious. The primary problem was how to submit the data to
the computer.
2. Keypunching was a possibility, but suffered from inherent
problems. CDH estimates in August 1980 for data entry costs
totaled $87,500 for 1.4 million inspections. Quality control
of the data would demand that each inspection be keypunched
twice, producing between two and three million records
annually. Because the keypunchers would be reading handwrit-
ten copy, there appeared to be additional room for error
beyond that intrinsic to keypunching.
12
-------
3. Using a machine readable form would eliminate keypunching.
The design of the form stipulates limits on the values which
can be coded. A higher rate of accuracy is thereby
promoted. The forms are traceable to the inspection station
and mechanic, even manually, if need be. No special equip-
ment is required by the inspection stations. The concept was
already familiar to the CDH staff because machine readable
forms had been used for several years in the Colorado
Mechanics' Training Program to record answers for the cer-
tification test and to gather demographic information (see
sections 2.5 and 3.2 of this document).
4. Purchasing special data collecting analyzers was considered
but would have placed undue economic burden on the inspection
stations. Although the automatically collected data would be
accurate, there would be a risk of the manually keyed,
vehicle specific data being inaccurate. Tracing or even
identifying the inaccuracies would be extremely difficult.
These units typically are self diagnostic, an advantage over
analyzers which do not collect data. Additionally, the
reviewing legislative body felt uncertain that the automated
analyzers ultimately would be available at the costs and
conditions stated by the interested manufacturers.
5. This system would be similar to that outlined under alternative
four, except the data tapes would not need to be manually
collected. The problems would be similar, but could be
compounded by the application of data transmission to a
decentralized program.
13
-------
Two other systems were proposed as combinations of type three and
four alternatives. One system attached to a standard analyzer, converted
the analog data to digital, and transmitted it to a computer. A major
concern centered around whether this "black box" would be compatible
with the variety of analyzers in use. The other system was simply a
keyboard on which all the data would be typed and thereby transmitted
to a computer. The typing seemed to represent one more step removed
from the actual data and could introduce an unusual amount of error.
Rather than attempt a hybrid system, the AIR program elected to use
machine readable forms.
2.5 Additional Investigations
Before the decision could be made to use machine read forms, the
topic needed further investigation. Optical scanning equipment can be
either Optical Character Readers (OCR), which decipher characters
on a page, or Optical Mark Readers (OMR), which simply detect the
absence or presence of marks, but not their shapes.
The COM Data Services Section, being better aware of the entire
scope of data needs for the Air Polution Control Division, had
explored the feasibility of optical character recognition (OCR) equip-
ment. Such a system actually ascertains which character is present in
a field rather than simply sensing whether a mark is present. OCR is
very versatile and could be used by other programs within the Health
Department. The writer is not limited to just a Number 2 pencil;
colored pen or pencil can be used. The equipment appeared to be more
reliable than mark sensors, and text could even be input into word pro-
cessing equipment elsewhere in the Department of Health. The primary
14
-------
problem with the AIR Program application lay in the OCR's speed: at a
read rate of 225 inspection forms per hour, it could not handle the
expected volume of data. Frequent operator intervention was also
required. Equipment costs can begin at $65,000 for this type of OCR
system. Another optical reader system which Colorado considered read
hand printed numerals. Unfortunately, alphabetic characters were not
as easily read, and Colorado vehicle licenses include both alphabetic
and numeric characters. Other drawbacks of using this system included
a high purchase price, a mechanics' training component which would be
needed to teach mechanics how to correctly write the digits, and
problems associated with the system's uniqueness: a back-up system
would not be available, and timely service could be unavailable.
After investigating the potential of this system, CDH concluded it
would not be feasible for reasons of reliability and cost.
The Colorado Emissions Mechanic Training Program provided a signi-
ficant opportunity to investigate a mark sensing system. Since
1977 Colorado State University, the contractor for the training
program, has been using a machine readable form to record demographic
information about the mechanics as well as the mechanics' answers for
the certification test. The forms underwent several design changes
the first several years. Because the forms are computer generated and
printed on campus, redesigning the form is not difficult. The form
can be printed on a variety of papers with dimensions from about
3" x 5" to 81£" x 11" or slightly larger. Red pre-printed text is in
an 80 character by 60 line area on an 81/2" x 11" page. The typeface is
a standard ten character per inch, six lines per vertical inch, all
upper case, numbers and selected symbols. An IBM 3881 Optical Mark
15
-------
Reader scans the forms for circles blackened with a Number 2 pencil in
selected areas of the form. It converts that data from printed form
to numbers recorded on a 7-track tape. The taped data can undergo
whatever computer analysis is necessary. The IBM 3881 can read an
average of 3,000 forms per hour, depending on the type of form.
Because mechanics complete this form when they take the qualification
test for AIR Program licensing, they are already familiar with the
concept of machine readable forms.
AIR Program staff was also familiar with Illinois' decentralized
truck safety data collection system. It has used machine readable
forms since 1975. In telephone discussions, an Illinois staff member
related that their program initially had a roughly 50% completion
rate, but that the rate had risen to a consistent 85% base rate on the
initial reading. Some errors are correctable, so that fewer than
1,000 forms out. of each batch are unusable. The Illinois program pro-
cesses 10,000 forns in each batch using an IBM 3881 optical mark
reader. More than 90% of the safety testers underwent training in
regional workshops. The Illinois program seems .to be the only other
program in the United States using machine readable forms in a garage
environment.
A data collection and handling system relying on coded forms and
a mark sensor seemed increasingly feasible for the Colorado AIR
Program. The ultimate decision would depend on two points:-
• the cost of the data system, and
• the availability and type of equipment in Colorado.
These two points were investigated by the CDH Data Services Section
for the AIR Program.
16
-------
A decentralized system was one possibility: several relatively
simple scanners distributed throughout the program area would read
x
inspection forms at decentralized collection sites. The data could be
transmitted over telephone lines to a centrally located computer for
analysis. The Scan-tron model scanner which was proposed costs
approximately $2,000 per unit. Unfortunately there was insufficient
time to investigate thoroughly the reliability of that equipment.
Were the AIR Program state-wide there also would be greater incentive
for a decentralized data collection system; the current program area
is within one and a half hour's driving time of Denver and currently
does not justify a more elaborate, and perhaps less reliable, com-
munications network. A centralized scanner system was decided upon.
Should CUH buy time on mark sensing equipment, or should it
purchase the equipment? Although there probably are enough applica-
tions within the Department to support the equipment, the decision was
made to buy services. The bids for services varied widely, but the
successful bidder proposed a package that was more cost effective than
purchasing the equipment. The compromise which offsets the lower
costs is the decreased availability of the equipment for programming
changes. Delays in changing source codes, for example, are to be
expected simply because the equipment is not conveniently on site.
Availability was the determining factor in deciding which equipment
to use. Data Services staff discovered that several sites were
available in the Denver area, and that most of them were public school
systems. Additionally, almost all of them used National Computer
Systems, Inc. (NCS) equipment, perhaps because of NCS' marketing
thrust. The NCS systems could cope with the types and volumes of AIR
17
-------
Program data, so the remaining point was cost. Denver Public Schools
(DPS), at 3800 York,.. approximately three miles from CDH offices,
entered into an agreement with CDH. Because DPS is also a governmental
agency, it could offer scanning services at $20 per hour, chargeable
by the minute. Its equipment is a NCS Sentry 7010 with limited computer
capability and a maximum read rate of 3,000 forms per hour. Practically
speaking, 2,000 forms per hour is a typical functional rate. A system
of this type can cost more than $85,000. A formal letter of agreement,
renewable annually, was made with DPS for the use of their equipment,
and that agreement has undergone minor modifications since it first
went into effect. The relationship between the AIR Program and DPS
seems cordial enough that minor problems in the system can be handled
without formal negotiations. Items to be considered in a formal
agreement with a contractor include:
• How many forms will be read in a given time period (e.g. 20,000
forms per week)?
• Will the contractor make arrangements to have forms read
elsewhere if his equipment goes down? Who pays the third party?
• Will the contractor guarantee use of the existing equipment over
several years or a recourse if the contractor changes equipment?
• What schedules will be kept by both parties?
• What accessibility for programming will be allowed?
• Can other services be provided (orienting forms, storage, etc.)?
• What will the costs be and what are the payment .schedules?
The Colorado AIR Program chose optical mark sensing equipment
because neither a less automated nor a more automated system seemed
feasible. Less automated systems could not cope with the data
volumes; more automated systems - specifically data collecting
18
-------
analyzers - put undue financial burdens on the inspection stations and
could prove less reliable than predicted. Other types of optical
scanning equipment were rejected because of high costs, lack of
availability, slow speeds, or poor reliability. Additionally, optical
mark sensing equipment was available at several sites and at
reasonable rates. A formal agreement with the contractor ended the
hunt for adequate equipment and moved the AIR Program data system into
implementation.
19
-------
3.0 IMPLEMENTATION OF DATA SYSTEM
3.1 Inspection Form
The inspection form is a crucial element of the data collection
process. The form currently in use by the Colorado AIR Program is the
product of months of development by AIR Program staff. Numerous
designs evolved over this period. Although some of the design changes
were necessitated by program modifications, most were initiated by the
AIR Program staff to clarify and simplify the form, making it more
efficient and insuring that it collected essential data. The develop-
mental process might have been shortened if the AIR Program staff had
been able to consult forms design experts much earlier. As it was,
the National Computer Systems, Inc. (NCS) class on forms design could
not be scheduled until June 1981, six months after forms design work
was begun. An NCS instructor taught the five day course in Denver for
the AIR Program at a cost of $2,200.
A concern of the program was to provide the motorist with a copy
of pertinent data for the inspected car. If the vehicle failed the
inspection, the owner could choose to have the vehicle adjusted other
than at the original inspection station and would need the inspection
information for a free re-test at the original station. A tear-off
part of the inspection form, kept in the glove compartment of the car,
was originally planned. The inspection procedure therefore required
entering selected data on both the top section and the bottom, tear-
off section of the form. Later, the motorist's form was converted
from a tear-off section of the first page to the entire third page of
a three page carbon form. Still, designing the front page so that it
was compatible with the last (third) page presented many complications.
21
-------
Changes initiated over the development of the form include:
• coding the inspection stations' and mechanics' numbers rather
than writing them in by hand
• avoiding double entries
• decreasing the amount of information to be coded
• collecting only total repair costs, not broken down by
labor and parts
• collecting rounded dollar amounts instead of dollars and
cents
• collecting number of cylinders, not cylinders and displace-
ment
• numbering the data blocks to help guide the mechanic in using
the form
• adding, then deleting areas where validation punches could be
used
• printing more explicit instructions on the form, and
• moving data blocks on the form
• for more prominence, and
• to follow the flow of procedures more closely rather than
grouping types of data.
The MCS scanner assumes alphabetic characters begin columns; a
program was written to restructure the license data, but the AIR staff
now knows the alphabetic characters in each column should precede the
numeric characters. This is an example of equipment-specific require-
ments which impact other aspects of the data system.
Based on the data required, the size of the form has become 8l/2" x
12", with a 5/8" strip at the top to hold the three pages and two car-
bon sheets together. The design of the form is a decision to which
both the Department of Revenue (CDR) and the Department of Health (CDH)
contribute. CDH handles the technical design considerations with MCS,
but the ordering and handling of the forms is a responsibility of CDR.
The printing costs vary from $.074 to $.096 per form, depending on the
quantity ordered. Additional mock-up costs are applicable (Appendix
F). They are ordered in quantities of 750,000 now that the program
is operating at full volume. Five weeks time from order to delivery
is typical. The forms are printed in color (July 1981 - December 1981
in blue; January 1982 and later in green) to aid distinction between
22
-------
the two forms. NCS prints the forms on its own "Trans-Optic"® bond
paper, which is guaranteed readable; other firms have informed the AIR
Program that they can print the forms at a lower cost, but no decision
has been made regarding changing from NCS printed forms. In Figure 1
are printed, in reduced format, the "blue form" used from July to
December, 1981 and the "green form" currently in use. The instructions
appear on the reverse of the third page of each inspection form. As
, i
an example of the types of changes which have occurred or which are
scheduled to occur, contrast the two forms in Figure 1 (full sized
photographs of the forms appear in Appendix D). Because of a con-
templated change in the standards, the next, revision of the inspection
form will eliminate the printed block with the state standards, thus
avoiding the obsolete form problem. The two empty blocks on the green
form have been reserved for the two-speed idle test data to be
gathered beginning in July 1982 for 1981 and newer vehicles. A copy
of this newest form also appears in Appendix D.
Although a complete redesign of the computer data capture program
took place with the change from the blue forms to the green, it was
necessitated by other factors (sections 4.1 and 4.2) than the form
revision alone. Even small changes on the form can require major
reprogramming because of the interrelations of data, data manipulation,
and equipment requirements.
Compromises on the form include some hand-written information
(collected as a last recourse for manual vehicle matching) and financial
data collected only to the dollar. The newest form asks the mechanic
to enter the actual cost, including cents, then code only the dollar
amount. This discourages rounding the dollar amounts. Compensation
factors are included in financial calculations where appropriate.
23
-------
Figure 1: FORMS CONTRAST
FIRST PAGE
Blue System
July - December 198)
Green Systeo
January 1982 - Current
ro
COLORADO DEPARTMENT OF HEALTH
-AIR" PROGRAM INSPECTION/READJUSTMENT
REPORT
VEHICLE IDENTIFICATION NUMBER
' MAKE
AMC O
AUDI O
AUHE©
AUST ©
IMWO
IUCK O
CADI O
CHEK O
CHEV
CHRY
DATS O
DODGO
FIAT
FOHO ©
BOND©
INTE O
IAGU O
JEEP O
LANC
UNC O
MA2DO
lEBZQ
MERCO
MG O
OLDS O
OPEL O
'O
PONT O
PORS Q
PUGT
RENA O
SAAB O
SUBAQ
TOY Q
TRIP
VOLK O
VOIV O
OTHRQ
® LICENSE PLATE
0©©©©0
©©©©©©
©©©©0®
0©©®®©
©©©©0©
®@©®®©
©©©©©©
©©©©©©
©©©©©©
I I I I I
©o©©o©
©©©©s©
©©©©©©
©0®®©®
©0©0©©
©0
O©
©
©
©
0
©
©
©
_©_
O®
®©
®©
©
©
©
©
0
©
©
©
©
©
©
©
®@
©S®©
©®O©
©®®©
®@©®
©©©©
0®©©
O©0©
©®@©
HI
0©Q®
©®®@
0©0©
®R5
©0
0®
0®
©S
Q RETE5T EMISSIONS LEVELS
o®
COSTS (IN DOLLARS)
I
o©
©
©
©
©
©
©
©O©
•tUWlKD REPAID
©e©
©®©
©©©
©®©
®e©
©00©©©©©©©
©©©©©©©©©©
©©©©©©©©©©
&
RETCST
VISUAL UUPHIIUi
FAILED
CATALniC CONVERTER O
FUEl RESTRICTOR O
AIR SYSTEM O
EXHAUST
SYSTEM INTEGRITY O
CERimCAnON ISSUED
COMPLIANCE O
ADJUSTMENT O
DENIED O
©®.o
0.0E,
®.®
©.©
000©
©oo©
© PASS ©
© FAIL ©
©0©
0®®
©0©
1 anil OM 1 km
at i«Y runn*
tin nte ud
ptrtonurf Out ioifiono
a nxordnci wilt
el to Coterido AIR
SIGUTUIE OF UCStSED EMISSIM HECHMIC
124693
© FIRST TUT
VISUAL IKSKCnOtt
CATALYTIC COKVERTER O
FUQ RESTRICTOR O
AIR SYSTEM Q
EXHAUST
SYSTEM INTEGRITY O
© HRST TOT EMISSIONS LfVEU
%co
1
0®L®
©©©
®J©
®L®
&S
<5X®
®G>
a®
Utti
ffmK
®®@®
Q©0©
®@®®
®®0
©0©
®e®
©0©
©0©
®@®
^0©
Q VOLUNTARY REPAIR
TUNE-UP O
CARBURETOR O
AIR CLEANER O
CHOKE O
OTHER O
@ HOME ADJUST
%CO p*aNC
H H_
PASSO
FAIl O
Q ADJUSTMENT ONLY
STATION
NUMBER
©®@®
©0O©
®®@®
®00®
®0©0
0©0®
®®®®
©00®
®®@®
®®@®
MECHAKIC
®®@®
©000
©00®
©0®®
0©©©
©®@®
©@®®
©00©
©00®
0®®®
PASS
FAIL
USE A NUMBER 2 PENCIL ONLY ON THIS FORM
FOLLOW INSTRUCTIONS ON REVERSE SIDE
THIS COPY TO DEPARTMENT OF HEALTH
PAGE1
COLORADO "AIR" PROGRAM REPORT
VEHICLE IDENTIFICATION NUMBER
VEH.
ADDRESS
AMC O
AUDI O
AUHE O
AllST O
BMW O
BUCK O
O
CBEK O
CHEV O
CHRY O
OATS O
DODO O
FIAT O
FORD O
HOND O
INTE O
JAGU O
JEEP O
LANC O
UNC O
MAZO O
MEBZ 0
MERCQ
O
OLDS O
OPEL O
FVFM O
PONT O
PORS O
PUGT O
RENA O
SAAB O
SUBA O
TOY O
TRIP O
VOU O
VOIV O
a O
© LICENSE PUTE
OOOOOO
©®®00©
O®®©0©
O©®0®©
©S®®®®
®00®©®
©0©©©®
0©©®©®
®©©©®®
©0©0©©
©0®®®®
00®®®®
©00OOO
©00©©©
©0©®©®
©0©O©©
©®®®@©
0©©®®0
©©®0®0
©s©®®®
©©©©©o
0©S®©O
©®
o©
©
©
©
0
©
©
0©
0©
©©
0©
©
0
©00©
®00©
©00©
0©©©
©00®
©00®
©00©
0®0®
®@®@
®©®0
Q FIRST TEST VISUAL
82 OR NEWER O
PASSED FAILED
© CATALYTIC CONVERTER ©
© FUEL RESTRICTOR ©
©AIR SYSTEM ©
O 81 OR OLDER
0 FIRST TEST EMISSIONS LEVELS
KCO
1 1
©©©
©0©
0©
0©
0©
0®
0©
0©
0©
0©
ihtfi
Inrif
ntfet
Itutufe
IMM.
P.. »C
©©©©
OOO©
©00©
©0©
©e©
®®®
©0©
©0©
0®®
©©©
'"© Flf O*
USE ONLY A NUMBER TWO 121 PENCIL ON THIS FORM
FOLLOW INSTRUCTIONS ON REVERSE SIDE
000®
0®®®
©0®©
0©®®
'5 ©
©O
®®
®0
©0
®®
®®
©O
®®
£1®.
0®
O©
©
©
©
Q CERTIFICATION ISSUED
82 Ofl NEWER Q
PASSfO FAIHO
© CATALYTIC CONVERTER ©
© FUEL fiESTfllCTOR Q
© AIR SYSTEM Q
Q B1 OR OIDEH
COMPLIANCE O
ADJUSTMENT O
DENIED Q
FINAL TEST VISUAL
©FINAL TEST EMISSIONS LEVELS
000
0®
®0
0©
0©
0©
0©
0®
0©
LL
©©©©
©ooo
©OS®
O®0
®s©
©0©
©©©
©©©
& PASS ©
°© FAIl ©"
564450
© "AIR" PROGRAM ADJUSTMENTS
©OQ®
©®®0
000©
0©®®
©®®0
©00®
©OO©
0©0©
©0O0
©00©
©0®®
©0©0
0®®®
00®®
MODEL ItU. 1111-
YEAR 1111 111.
eoiM T.O
HCIip>l im
1 I emifr tbil I hurt BtrforanJ thii inspection in »
with tht talM ltd gtudilmn gf tbi Catend* AIR Pragitm.
SICNATUIK (U LICFJIUD EWSSTCH MICHAIIC
THIS COPY TO BE PICKED UP 8Y DEPARTMENT OF REVENUE
-------
FIRST PAGE
ro
en
Blue System
July - December 1981
Not collected.
Title: Department of Health "AIR" Program
Inspection/Readjustment Report.
Alternate coding columns shaded for ease of use.
Data block 16 groups all dollar costs.
Instruction say "Cost (in dollars)" and only dollar
blocks are available. Rounded dollars could be
given. Repair costs are collected for both parts
and labor.
Data block 9: attached to block 10, producing a
complicated looking box. Circle coded only if
status is "fail." On which models the visual
inspection is to be conducted is not specified.
Data block 10 and 12: HC and CO pass/fail circles
not boxed.
Data blocks for adjustment and retest (11, 12, 14)
scattered on page. Second test labeled "retest."
Block 13: home adjustment data captured.
Emissions standards printed lower left. Signature
block in middle of page.
Instructions: "Use a number 2 pencil on Iy on this
form...This copy to Department of Health."
Green System
January 1982 - Current
Certificate (sticker) number written in allows
matching of sticker, vehicle, and inspection
form if necessary. Vehicle owner's name and
address recorded for manual identification if
necessary.
Title: Colorado "AIR" Program Report.
All mandatory data blocks for the first Inspection
shaded to focus attention on them, and alternate
coding columns shaded darker.
Data block 9 is inspection cost, grouped with
other inspection activities, not other costs.
Mechanic writes in dollars and cents, then codes
only dollars.
Adjustment costs appear in block 12 with other
adjustment data. Repair costs (total only and
requested specifically for 1981 and newer models)
are in block 13.
Exhaust system integrity section as separate block and
more clearly labeled. More complete instructions,
Including mechanic identifying whether 81 or older/
82 or newer vehicle. Pass and fail options avail-
able, not just fail, and Iabeled as P and F within
coding circles.
Pass/fail circles included in data blocks and more
clearly labeled.
Adjustment data grouped (12, 13, 14, 15). Second
test now called "final test."
Block eliminated.
Emissions standards printed lower right, over
bottom right signature block. Space left on form
for data blocks to be added as needed.
Instructions rewritten for accuracy: "Use only
a number two (2) pencil on this form...This copy
to be picked up by Department of Revenue."
-------
SECOND PAGE
Blue System
July - December 1981
Green Systeai
January 1982 - Current
ro
en
NCft Tram-Optic B1C
COLORADO DEPARTMENT OF HEALTH
-AIR" PROGRAM INSPECnON/READJUSTMENT
REPORT
0
VEHICLE momnttTia UMBER
' MAKE
AMC O
AUDI
AUHEQ
AUST O
IMVlO
IUCX O
CADI_O
CHEK'O
CHEV
CHRY O
DATS O
OODGO
FIAT O
FORD O
BOND©
INTE ©
JAGU ©
JEEP ©
LASC ©
LIKC O
MAZDO
MEBZO
MERC©
O
OLDS ©
OPEL ©
aYM
PONT
POHS O
PU6T O
RENA O
SAAB O
SUBA O
TOY O
TRIP O
VOIK ©
VOLV ©
OTHRQ
© LICENSE PLATE
©OS©©©
©®©©©®
®®©®©©
©©©©©©
®®©®s©
©©©©©©
©®©@®®
©©©©0®
©0©©©0
0®©®©®
©®©®©®
©©©©©©
©®©©©®
©©©©©©
©o®©©©
©©©©S©
©©
©
©
©
©
©
0©
©®
©®
©
0©©0
0®©0
©©©©
©®0®
©©©©
©0©Q
®©®0
®@®0
©0©©
©s&
o®
®©
®®
Q©
®©
®©
0®
©®
@®
®S
®©
©0
© CERTIFICATION ISSUED
@ HFJtST EMISSIONS LEVELS
COMPLIANCE ©
ADJUSTMENT ©
DEWED ©
l~cT1
©0X3
©0
I
@ COSTS (IN DOLLAR!)
IBIP.
Z
©©
©
©
©
©
©
©
©
AU.
®©
®©
©
©
©
©
©
©
©
REaillRID REPAIR
LABOR
®®@
©e©
®ffi®
®@®
®S®
©S®
0®0
©©©
©®®
©®®
PARTS
I©©Q©©0©©0©
ks>®©©®©®®0®
|©®0©©©©©0©
0 RETEST
VISUAL INSPECTUM
IMS
CATALYTIC CONVERTER O
FUEL RESTRICTOR O
AIR SYSTEM Q
EXHAUST
SYSTEM INTEGRITY O
©.©
©®
©L®
©L®
(HtH.
©0©
©®0
©(?)(•)
© PASS ©
© FAIL ®
0 1 entity Uul 1 km pifomd tfcii inpwtbui
•ad lay rtfidnd irfJMtMmts at McordaftM with
tbi nta nd fuUma tl fe Cikmdo Alfl
SIGNATURE OF LICENSED EMISSION MECHANIC
124693
® FIRST TEIT EMISSION! LEVELS
n»«t
FIRST TEST
FAILED
CATALYTIC CONVERTER O
FUEl RESTRICTOR Q
AIR SYSTEM Q
EXHAUST
SYSTEM IHTEMITY Q
T
©©©
©0©
©©
©.©
®i®
®.®
a®
©00©
0©0®
©0©
©0©
©0©
©0©
©0©
Q VOLUNTARY REPAIR
TUNE-UP
CARBURETOR
AIR CLEANER
CHOKE
OTHER
[wtcitrl
PASS©
FAIL O
Q ADJUSTMENT ONLY
©®@®
©©©©
©©©©
©©®©
©©©©
®©®0
©©0©
®@®©
®®S©
©®®©
0©0©
©PASS ©
® FAIL ©
THIS COPY TO BE FILED AT AIR PROGRAM STATION
COLORADO "AIR" PROGRAM REPORT
tiCS Trim-Optic BIO 3204
VEHICLE inEHTIHUTHM NUMBER
AMC O
AUDI O
AUHE ©
AUST O
BMW O
BUCK O
CAOI O
CHEK O
CHEV O
CHRY ©
OATS O
OODG©
HAT O
FORD O
HOND©
IIHE O
JAGU O
JEEP O
LANE O
LUIC O
MAZOO
MEBZO
MERC©
MB O
OU1S O
OPEl O
PtYMO
PONTO
PORS O
TOT O
RENA O
SAAB O
SUBA O
O
TRIP O
VOIK O
VOLV O
OTHR C
© LICENSE PLAH
©®0©©©
®©®©©®
©©©©S®
®©®®©©
®®®®@®
®®®®®®
®®0®©©
©©©00©
©©©©©©
©©©©©©
©©©©©©
0©©©®©
®®
o©
©
®
©
®
®
®
®
S
©©
0®
©©
©
©
O©©©
®©®©
©©S©
©S©©
©00©
0®®O
0©®©
0®®®
©©©•©
®®s®
®00®
©S
0©
©0
©
©
®
®
®
® CERTIFICATION ISSUED
SI OB NEWEB O
PASStD FAIliO
CATALYTIC CDNVERTEB ©
© FUEl RESTRICTOR ©
© AIR SYSTEM ©
Q Bl OR OLDER
COMPLIANCE ©
ADJUSTMENT Q
DENIED Q
FINAL TEST VISUAL
® FINAL TEST EMISSIONS LEVELS
©0®
©0©
0®
©©
©®
©©
©©
®®
©<•)
®©@®
©©©©
©©©©
©S®
^ X !•
®
HRST TEST VISUAL
82 OR NEWER O
PAiMD FAILED
® CATALYTIC CONVERTER ©
© FUEL RESTHICTOfl ®
© Alfl SYSTEM . ©
O Bl OR OLOER
FIRST TEST EMISSIONS LEVELS
SCO
TT
0©G
®O
®0
®0
®O
®®
®©
©®
tO PASS ©
© FAIL ©
©0®
®® ©©©©©©©©
©©©®©©©©0®
©©©©©©000©
564450
•AIR- PROGRAM ADJUSTMENTS
O00Q
©©@®
®©©®
©©©©
®©Q©
©S®®
S©®®
©00©
0®®®
MODEL
YEAR
C0|%)
HC(pp»]
1110-
1171
1.0
1100
(& letnrfTtk.il
EMISSIONS STANDARDS
1171- till- 1177-
1174 1tTI 1171
1.0 S.I 3S
1100 m soo
b«t pttfwwri tto wsptttittD in ttemtt
at vuMiMt a Ib. Col.rwJ* AIR Progr
1I7U
NEWER
1.0
400
net
It&IATUU Of UCUSIO iHUSIOlW MtCHJUIIC
THIS COPY TO BE FILED AT 'AIR- PROGRAM STATION
-------
SECOND PAGE
Blue System
July - December 1981
Identical to page 1,
Green System
January 1982 - Current
Identical except exhaust system block is converted
to ten lines for remarks.
ro
"T. i
-------
THIRD PAGE
Blue System
July - December 1981
Green System
January 1982 - Current
ro
CO
COLORADO DEPARTMENT OF HEALTH
"AIR" PROGRAM INSPECTION/READJUSTMENT
REPORT
O
VEHICLE IDENTIFICATION NUMBER
124693
COLORADO "AIR" PROGRAM REPORT
AMC O
AUDI O
AUHE O
AUST O
BMWQ
BUCK O
CADI ©
CHEK O
CHEV O
CHRY O
OATS O
OOOGO
FIAT O
FORD O
HONDO
INTE O
JAGU O
JEEP O
LANC O
LINE O
M«OQ
MEBZQ
MEflC O
MG O
ILDS O
OPEL O
PLYM
PONT O
0 LICENSE PLATE
fflffl
LLL
IMPORTANT INFORMATION TO VEHICLE OWNERS
1. You must keep both halves of this form to present to
inspector at the time of vehicle safety inspection.
2. You must have both halves of this form to transfer to the oew
owner should you sell this vehicle.
3. YOU MUST HAVE A CERTIFICATION OF EMISSIONS CONTROL
BEFORE YOU SELL THIS VEHICLE.
4. KEEP THIS CERTIFICATION WITH YOUR REGISTRATION.
PLEASE
BEAD
WARMING
BELOW:
COSTS {IN DOLLABS)
WARNING:
If the retest visual inspection portion of this form |boi 11)
indicates that this vehicle is missing any emissions control
equipment, you may be required to have this equipment
installed before you receive your next emissions inspection
and/or safety inspection. Please contact a licensed emis-
sions mechanic or the Department of Health for specific
information.
ion TO VIH1CIE ITWHtB
RFTEST
VISUAL IHSPtCnON
FAILED
CATALYTIC CONVEHTER Q
FUEL RESTRICTOR O
AIR SYSTEM Q
EXHAUST
SYSTEM INTEGRITY Q
I certify that I have pBriormsd this inspection
end any required adjuitmenli in accordance with
the rules and guioelii»s of the Colo,ado AIR
Program.
SIGNATURE OF LICENSED EMISSION MECHANIC
124693
© FIRST TEST
VISUAL INSPECTION
FAILEO
CATALYTIC CONVERTER O
FUEL RESTRICTOR O
AIR SYSTEM Q
EXHAUST
SYSTEM INTEGRITY O
© FIRST TEST EMISSIONS LEVELS
% CO
1
<2)©G?<3G?e?QQ<=?<3
©©©©©©©©©©
©©
C«np*rt
tbttl
Intli
MtN
(w aofiil
ftlr
hi*
tllttil,
ffm HC
@@©0
00O©
©0©©
©0©
©0©
©0©
00©
©00
©00
©00
© VOLUNTARY REPAIR
TUNE-UP O
CARBURETOR Q
AIR CLEANER Q
CHOKE O
OTHER O
(specify)
Q HOME ADJUST
!**• Nslt to
ViMctTDmif)
% CO MM HC
| |
PASS©
FAIL O
Q ADJUSTMENT ONLY
STATION
•UMBER
IMPORTANT -
It will ilfow yog
iiupttliod 11 thf
ttilwi within tm
ficli iW firn imp
FOR ANY AODITI
CONTACT:
COLORADO DEPA
Ottwtf; 3
Avrwi: 3
Fi. CoHint: 2
MECHANIC
NUMBER
1 1 1
Mtf THIS FORM
a havt >nt !r« it
djmT^rMkklt
cirari.
NAL INFORMATION
TMF.NT OF HEALTH
D-4 1 80
M-4135
t-5324
PASS
FAIL
nuTiiiriTi »n
© VEHICLE IDENTIFICATION NUMBER
ft AUTO
MAKE
AMC O
AUDI O
AUHE O
AUST O
BMW O
BUCK O
CADI O
CHEK O
CHEV O
CHRy O
DATS O
ooocO
FIAT O
FOBO O
HONDO
IKTl O
JAGU O
JEEP 0
LANC O
UKC O
MAZOO
MEraO
MERCO
MG O
OLDS O
OPU O
PITM O
PONT O
POBS O
PUGT O
RENA O
SAABO
SUBAO
TOY O
TRIP O
VOLK O
VOLV O
OTHRO
nu nw»« •<•>.
564450
rrrv.
©„««..,.„ © DATE OF TEST 1 <7> STATION (!)•<»««;• ffl- (i) 5
III 1 II 1 II 1 1 1 1 1 1 1
©
IHKCTIQN
COST
•rrr
®
©CA
©FU
©AIR
HRST TEST VISUAL
® CERTIFICATION ISSUED
COMPLIANCE O
ADJUSTMENT O
DENIED O
(J7) FINAL TEST VISUAL
BZ Ofl NEWER O
fASSJO FAILED
© CATALYTIC CONVERTER ©
© FUEL RESTRICTOR ©
© AIR SYSTEM ©
OBI OR OLDER
%CO
I.!
©©©
0©.©
®®
©•©
©.©
0®
©•©
®e
W-HC
1 1
£S"" ©00©
ITS ®®©0
itMferii ©©©
El ®®e
©©©
© PASS ©
"© FAIL ©"C
564450
82 OR NEWER Q O "*IR" PROGRAM ADJUSTMENTS
FAIIED STATION MECHANIC'S ADJUSTMENT
AIYTIC CONVERTER © mint* NUMBER COST
L RESTRICTOR © 1 * 1* 1* I*
SYSTEM (?) 1 1 . 1
© EMISSIONS
REPAIR COST
111 OK NEWER)
•rr.T
0. ro« OIK. ,..,.."!•.!• "r.T"! .,
© FIBST Tin EMISSIONS LEVELS
©©.©
0.®
©0
0©
0©
• <•>•©
0®
©000
®@®
0®0
©0©
©0©
©0©
®
©
PASS
FAIL
THIS COPY RETAINED BY VEHICLE OWNER
II mitt *Hm T*> » taw «M frit »-
i*if»ttm n tk« irifiHl Mtptttim,
itilim wilfcia un ttjt it itta vikicti
l*il* it* lint i
.
t ".« min n i«ni m d ita Mln.»i c.^tu
) Fm tat!) t*i triHd inl III Mitwm hull i
FOR »*t ADDiriDML HIFOH«l*TrOIt. COKTACT OH OF TMt FOUOIr>l« I
TECHIIUL CIMTERI:
Otwnr Mini «•« • 1141 E. ttrtti »w . *»•'• 1W413S
F»1 CilliM Af« - 411 • Ollifi An.. Fl Cillw 221 S«»
tMt. tfiiftt *TM - 14U I. TtjM H . C*l* t»r»D I»-I111
OR THE COtOMBO DCPARTMEUT OF REVEMIt IH-kitl
EMISSIONS STANDARDS
C0(%) 7.0
HC(ppm) 1200
© l«rtifTthit tun
wiA Iln mill tnd
tW» wsptetion n tccoTdinu
at tbt Coterido AIR ProgiMi.
SICNATURE Of IICCNSEO EMISS.OHS HECHAIIIC
THIS COFT TO BE RETAINED BT VEHICLE OWNER
-------
THIRD PAGE
Blue System
July - December 1981
Information for the vehicle owner appears in four
sections on the form. CDH I isted as source of
further Information, along with just phone numbers
of Technical Centers.
Block 19: punch blocks to verify authenticity of
the Inspections, at the right edge.
Green System
January 1982 - Current
Information for the vehicle owner consolidated
Into two blocks, both of which are in the lower
right corner. CDR IIsted as source of further
information, along with phone numbers and
addresses of Technical Centers.
Punch blocks eliminated because the form is not a
control led document.
ro
10
-------
INSTRUCTIONS (REVERSE OF THIRD PAGE)
Blue System
July - December 1981
Green System
January 1982 - Current
to
O
INSTRUCTIONS TO LICENSED EMISSIONS MECHANICS
(ONLY a Licensed Emissions Mechanic may perform this inspection.)
NOTICE
The first copy (top) of this form will be scanned electronically.
Please keep the top copy free of dirt or grease.
Do NOT fold, staple, spindle, or mutilate top copy.
Do NOT make ANY marks on the top copy other than specified in directions.
You may make notes on the second or third copies ONLY.
Use ONLY a NUMBER TWO pencil on this form.
Where boxes are provided put one and only one number or teller in each box. if there are more boxes
than numbers or letters put enough zeroes in front {to the left) to fill up the extra boxes. For example,
if the emission levels are 4.8% CO and 750 ppm HC:
DO NOTE
Except for Item 0 (Vehicle Identification Number), after the boxes are marked, fill in the circle under
each box which has a letter or number that matches the letter or number in the box above. There
should be just one circle filled in under each box.
fIRST INSPECTION
A. Complete items O through 0.
B. Do the first test visual and emissions inspection © and G, (mark PASS or FAIL for CO and HQ.
C. If (he vehicle passes bath the visual and emissions inspection, mark "Compliance" ©. sign the form G. punch | jflC |
G. and "Date ol Test" on pages 2 and 3, and give both halves of page 3 lo the vehicle owner.
D. If the vehicle fails the visual or emissions inspection and the adjustments or repairs are done at your station, complete
item G and go on to Step G.
If the vehicle owner wants the work done elsewhere, give him/her the bottom pan of page 3 to be filled in by the person
who does the work and returned within 10 days for a free reinspection.
REINSPECTION
E. If the vehicle was adjusted or repaired by a Licensed Emissions Mechanic, copy the station and mechanic's license num-
bers from the "Adjustment Verification' form under item G. attach one copy of the "Adjustment Verification' (arm to page
2, and go on to Step G.
F. If the vehicle was adjusted or repaired by a non-licensed person, complete item G. If the vehicle passes, go on lo Step
G. It it fails, make the adjustments or repairs required and go on lo Step G or mark "Denied" ©. sign the form G.
punch CD G. and "Date of Test" on pages 2 and 3. and give both halves of page 3 to the vehicle owner.
G. If the vehicle failed the first test visual, do the retest visual inspection G. it it passed the first test visual, go on to step H,
H. Do the retest emissions inspection G. (mark PASS or FAIL for CO and HC).
I. Record the amount charged for the inspection, the adjustments (if performed), and the labor and parts costs for required
repairs G.
J. Record any voluntary repairs G.
K. If the vehicle passes both the visual and the emissions retests. mark "Compliance" G. sign the form G. punch [COT]
G. and "Date ot Test" on pages 2 and 3, and give both halves of page 3 to the vehicle owner.
L If the vehicle passes the visual retest but fails the emissions retest {and. for 1981's and newer. $100.00 was spent on
emissions repairs) mark 'Adjustment" G. sign the form G. punch | COA | G, and "Date ol Test' on pages 2 and
3. and give both halves of page 3 to the vehicle owner.
M. It the vehicle fails the visual retest (or, lor 1981'j and newer, fails the emissions retest and less than $100 was spent
on emissions repairs) mark "Denied" ©. sign the torm G. punch (~CO | G, and 'Date of Test' on pages 2 and
3, and give both halves of page 3 to the vehicle owner.
• Thi (ml copy (top) at ihii (arm will b* ttonntd iltctronicilhj.
• Plttst kttp tht lop copy Irtt il din or grm*
* Do NOT (old. staplt. ipindlt. at miniltti top copy.
• Do NOT miti ANY rnarkj on tht tgp copy othn thin specified in direction*.
• You miy ntkt natit on tht ucont) or third copin ONU.
• U» ONLY i NUMBER TWO ptncil on IRII larm
• Whin torn irt providtd pui ont and only oni numbtr o< lonti in tach
boi. il inert •'( matt bom thin numbiri 01 It tier» put enough itroti in
tfpni (to tht lift) to (ill up tht inri bom For tiamplt. il (hi uniitian
Itvcli tit 4 §\ CO and 750 ppm HC:
ENTER
INSTRUCTIONS TO LICENSED EMISSIONS MECHANICS
(ONLY a Licensed Emission Mechanic may perform this inspection.)
NOTICE
• Eictpt fm Htm 0 (Vthklt Idtntif ication Number), alltt the bo in in
marktd. Ml in tht circlt under mh boi which hai a Itner 01 number thai
mitfhts the liner or nutntui in tht bet above. Thtrt thoutd b* jusl oni cirdi
lilted in undu nch boi.
FIRST INSPECTION
PRINT • Name and iddress of vehicle owner at top of this form
ITEM
O Vehicle Identification Number - Write in VIN number.
0 Auto Make • Fill in circle for auto make.
0 License Hale - Enter license plate number and fill in circles (if
temporary tag leave blank, if more than six characters use first tin).
0 Date of Test • Enter month, day and year and fill in circles.
0 Station Number - Enter your nation's license number and fill in
circles.
0 Mechanic's Number • Enter your emissions mechanic's license
number and till in circles.
0 Model Year - Enter last two digits of vehicle's model year and fill in
circles (example: 1972 enter 72).
0 Na. ot Cvi • Enter number of cylinder* in engine and fill in circles
(if rotary engine enter number of rotors).
0 Inspection Cost • Enter cost of inspection in dollars and cents
(maximum $10.00} and (ill in circles under dollar portion only.
© First Test Visual - For 1981 and older vehicles, fill in "81 OR
OLDER' circle (no visual inspection needed).
For 1982 and newer vehicles, fill in '82 OR
NEWER" circle, complete first visual inspection
and fill in a pass or fail circle for each emissions
control system,
O first Test Emissions Levels • Complete first emissions inspection,
enter readings and fin in circles. Compare readings to State standards
and till in pass or (ail circles for CO and for KC.
NOTE; THIS COMPLETES FIRST INSPECTION. IF VEHICLE HAS
PASSED ALL REQUIREMENTS. FILL IN 'COMPLIANCE' CIRCLE IN
ITEM G AND SIGN ITEM @ . ENTER CERTIFICATE Of EMISSIONS
CONTROL NUMBER AT TOP OF FORM. GIVE THIRD COPY TO
VEHICLE OWNER AND AFFIX CERTIFICATE OF EMISSIONS
CONTROL STICKER TO THE WINDSHIELD AS SPECIFIED.
VEHICLES FAILING FIRST INSPECTION
ITEM
G "AIR" Program Adjustments - If adjustments were made by a
licensed emissions mechanic, enter license numbers of station and
mechanic where adjustments were made and cost of adjustments in
dollars and cents (maximum $15.00). Fit! in circles under station
number, mechanic's number and dollar portion of adjustment cost.
NOTE: If adjustments were made by another licensed emissions
mechanic, copy this information from "Adjustment Verification Form".
G Emissions Repair Cost - For 1981 and newer vehicles, enter cost of
emissions related repairs in dollars and cents and fill in circles
under dollar portion only.
G Final Ten Visual - For 1981 and older vehicles, fill in "81 OR
OLDER* circle (no visual inspection ntedtd).
For 1982 and newer vehicles, fill in '82 OR
NEWER" circle, complete linal visual inspection
and fill in a pass or fail circle for each emissons
control system.
G final Ten Emissions Levels • Complete final emissions inspection.
enter readings and fill in circles. NOTE: II adjustments were made
by another licensed emissions mechanic, copy readings from
"Adjustment Verification Form'. Compare readings to State standards
and fill in pass or fail circles for CO and for HC.
njssued • Fill in circle next to:
COMPLIANCE rl torn (R> passed end item © is within Slate
standards.
ADJUSTMENT if item © passed, item G Wed. adjustments
done by licensed mechanic and, tor 1981 and
newer vehicles, al leasl $100 spent on
emissions related repairs.
DENIED if hem G (illed « """ G tail*d arid ')
adjustments not done by licensed mechanic or
2) for 1981 and newer vehicles, less than
$100 spent on emissions repairs.
G Signature ol licensed Emissions Mechanic • After completing ihis
form, sign it. enter number of the certificate issued (or 'none" it
certification denied) at top of form and give third copy to vehicle
owner. It a certification is issued, affii certificate o( emissions
control nicker lo the windshield as specilnri .
-------
INSTRUCTIONS (REVERSE OF THIRD PAGE)
Blue System
July - December 1981
Instructions grouped by 1) form, 2) first
inspection, and 3) re-inspection. Printed across
the page, alphabetically labeled in steps.
Green System
January 1982 - Current
Instructions organized by 1) form, 2) first
inspection, and 3) vehicles failing first
inspection. Boxed by topics. Instructions ordered
by data block numbers for easier reference. Each
block has instructions.
-------
3.2 Hechanics' Training
Mechanics wishing to be licensed to conduct AIR inspections and
adjustments must first be qualified. "Qualification" consists of
passing a written test (25 technical questions, 25 rules and regula-
tions questions) and a performance test. The mechanics may take the
qualification tests free of charge by appointment at one of the three
Vehicle Emissions Technical Centers (VETO in the program area.
However, the mechanics are encouraged to enroll in a 16-hour training
course before attempting the tests. The cost of the course is $25.00,
which covers the instructors' and facilities' fees and mailing. The
course consists of four classes (usually Tuesday and Thursday, 6 p.m.
to 10 p.m., two consecutive weeks) conducted by a trained instructor.
The text book for the course has been developed under contract by the
Colorado State University project staff over a span of several years,
with revisions and additions being nade as necessary. One unit of
that training manual is devoted to completing the machine read inspec-
tion fora. The unit appears in its entirety in Appendix C. A mock-up
of an inspection form was included in the manual even before the first
form was finalized or in use. That unit has been revised regularly.
One hour of the third class and one hour of the fourth class include
instructions on the use of the inspection form, and the performance
test requires the mechanic to complete an actual inspection form.
Seven questions in the 85-item computerized bank of rules and regula-
tions questions relate to the inspection form. A mechanic not
electing to take the emissions course may obtain a free training
manual at a Vehicle Emissions Technical Center (VETO but receives no
32
-------
additional instruction on the form. The machine read answer sheet on
which the mechanic marks his responses to the written qualification
test reinforces some of the basics of completing optically read forms:
• use a Number 2 pencil,
• mark only one circle in each row or column,
• fill each circle completely and neatly,
• complete all items, and
• do not fold or staple the form or get it wet or dirty.
It is difficult to isolate from the other training topics the
costs of training mechanics to complete the inspection form. Because
two of the sixteen hours of instruction are devoted to the form, one
eighth of the 325.00 fee could be considered part of the costs
($3.13/mechanic). The training manual is printed under contract by
Colorado State University at a cost of approximately $4.90 each. The
unit on inspection forms comprises roughly 10% of that document, or
$.50. The development of the unit has either been completed by the
Colorado State University project, staff under contract to the
Department of Health or, in 1981-82, by the AIR Program staff. It
entails about 80 man hours and an estimated $100.00 worth of black and
white graphics taken from the actual inspection form. The initial
development represents an investment of perhaps 31,500; minor revi-
sions are considerably less. Twelve color slides pertaining to the
inspection form were developed for the class instructors, at a cost of
about $7.60/slide. Each slide consists of a photograph of an actual
inspection form covered with a sheet of yellow transparent film over
all except the data block being discussed. These slides are repro-
duced at Colorado State University at a cost of $.38 each in quan-
tities of 25 or more.
33
-------
f
As difficult as it is to isolate the expense of training mechanics
to complete the inspection form, some identified costs include:
per mechanic: $ 3.13 class
.50 text
$ 3.63 (does not include testing
or record-keeping)
development: $1,500 initial text development
850 major revision to text
91 (12 x $7.60) slide development
140 (12 x $.38 x 30 sets) slide production
$2,581
There are portions of other training program costs (proctoring the
tests, for example) which could be attributed to training for the
inspection form, but these costs are minor and would be incurred
whether or not the inspection form were included in the training.
3.3 Data Input
Data input is the process by which the inspection forms are obtained
by the licensed mechanics, retrieved by the Department of Revenue
(CDR), data captured fron the form and analyzed. Figure 2 diagrams
this process.
The Department of Revenue (CDR) is responsible for ordering the
forms from NCS. The forms are stored by CDR and obtained by the
mechanics fron the CDR office in Denver when they purchase the window
stickers. One form is provided with each sticker. CDR inspectors
also carry forms in their vehicles in case a station should need any.
Although the forms are consecutively numbered, they are not controlled
documents, and no security is required, either by CDR or the stations.
There is no charge to the inspection station for the forms. Identification
is required when the forms are obtained only if stickers are obtained
at the sane time.
34
-------
Figure 2: DATA INPUT
JANUARY 1, 1982 TO PRESENT
( START )
CDR
Receives Torms from
NCS and stores them
Inspecting mechanics
tJbtaln forms
from CDR
Inspecting mechanics
Use forms
Inspecting mechanics
Forms separated
Page 1 for
analysis
Page 2 retained
by station
Page 3 to
Vehicle Owner
CDR inspectors
Forms picked up in
labeled envelopes
every 50-60 days
CDR
DelIvered to uenver
pub Iic schools
reader week Iy
/CDH\
/Manual qual ItA
o
to
CO
OJ
CDH
'Manual qualIt
( control check >
\ and page /
\ orlentatlon /
Yes
CDH
CDRInspectors
notified and
asked to relay
to mechanics
Scanner stamps
each form read
with a sequential
number
-------
After the inspection process has been completed by the mechanic,
the three sheets of the form are separated at the top perforations.
The third page is given to the motorist; the second page is retained
by the inspection station for its records. Inspection stations are
told to store their copies for one year. Page one is picked up during
the regular audit by the Department of Revenue (CDR) inspector
assigned to the station. Audits are mandated to occur no less fre-
quently than every 60 days, but in practice the CDR inspectors have
been auditing the stations every 30 days. Any questions the mechanic
might have regarding the form can be answered by the CDR inspector,
but the inspector is under no obligation to check the forms collected
from the station. The collected forms are returned to the Department
of Revenue (CDR) office in Denver and are delivered weekly to the
Denver Public Schools (DPS) scanning site. Nadine Quig.ley of the
Department of Health (CDH) Mobile Sources staff assumes responsibility
for data handling from this point until a data tape is provided to CDH
Data Services. At the scanning site a Department of Health (CDH)
Mobile Sources data clerk conducts a manual quality control check (for
obvious tears, etc.) and orients each page for reading by the NCS
Sentry 7010 optical mark reader. When the projected data volume of
20,000 forms per week is achieved, the volume will dictate that the
quality check preceding machine reading be perfunctory. Any problems
with the completion or condition of the inspection forms are reported
to the Department of Revenue (CDR) inspectors weekly on a notification
slip. They relay the information to the inspection stations, sign to
verify the station has been informed, and return the slip to the
36
-------
Department of Health (CDH). Serious and consistent problems can also
be relayed to the mechanics' training program as a suggested revision
to the instruction on conpleting the inspection forms.
Significant differences between the "blue" and "green systems" are
apparent at the "forms scanned" stage. The data input system designed
for the July-December ("blue") phase concentrated editing at the
Sentry 7010. The reader attempted to scan each form. Forms which could
not pass through the reader were ejected into a bin. Forms which phy-
sically could be scanned were then read and problems identified. A
"problem" embraced data missing or not readable, data out of range,
and conflicting data. The forms were shunted into the rejection bin
if a serious enough problem was found. The Department of Health (Cf)ll)
had established (and could alter) criteria by which the scanner judged
the severity of the errors. A numerical coding system identified the
types of errors found, and the appropriate code was stamped by the
machine on the back of the rejected form. Forms judged by the scanner
to include accurate data were scanned, stamped with a sequential
number, and deposited into a separate bin. They were then ready for
boxing and storage by sequential number. Although very "efficient"
data vjere captured with this systen, data useful for some purposes
were lost. Data containing fatal errors were not captured on the com-
puter tape; a check to ascertain if forms with errors were represen-
tative of the entire data base could only be accomplished manually.
The limited computer capability of the optical mark reader was also
taxed with the error identification codes, so error categorization
failed to be as complete as hoped.
37
-------
Redesigning the data handling system for the full-volume ("green")
phase of the AIR Program provided an opportunity to eliminate several
of the problems noted above. The editing was transferred from the
mark reader to the computer. Data are captured from every form which
can possibly pass through the scanner, and the computer assesses and
notes the accuracy of the data. In this fashion the most complete
data bank possible is created. A computer check can then be made to
test whether errors are randomly distributed. This procedure captures
a greater amount of data, and as a result of a sophisticated error
identification system (section 4.2), even data with questionable errors
are usable for selected purposes.
There exists an underlying question at this point which could perhaps
be titled "The Ethics of Data Capture." Should a form which obviously
contains errors undergo correction? Or is it to be preserved as is
because it represents a significant statistic? If it is to be corrected,
can one be sure the alterations are correct? Colorado has not elected
to correct invalid forms. The volumes and types of errors are indica-
tions of the "state of the AIR Program" and are significant in themselves.
The inspection forms are technically "source documents" and origi-
nally had to be retained for at least two years. The Department of
Health (CDH) feels that a storage time of 90 days or until CDH is con-
fident of the data tape accuracy is sufficient. Provided that the data
tape is accurate, this decision is reasonable: to physically locate an
inspection form would probably require three days' search time. To
find the computer record for that form takes a matter of minutes.
Handwritten information on the form (owner's name and address, VIN,
sticker number) is not essential for the automated data system.
38
-------
One week is allowed for the scanning process. A new batch is
begun weekly. During the July-December ("blue") phase, the batch size
averaged 3,000-5,000 forms. Three months into the "green system," the
volume had developed to more than 20,000 forms per week. The pro-
jected volume was 20,000 forms per week. Without an efficient data
handling system, a program of this magnitude would be floundering in a
sea of forms and data.
3.4 Readability
Readability is one component of the overall "completion rate." If
the nark, sensor accepts the form for scanning, it is classified as
readable. The other component of the completion rate, the integrity
of the scanned data, is discussed in sections 4.1 and 4.2 of this
report. If the form jams in the machine, if the form is torn so that
the sensor cannot scan it, or if critical form marks (bias bar, skunk
marks, timing marks, etc.) are marred, the scanner rejects the form.
Comnon problens which cause forms to be rejected include the
following.
• The service writer staples sheets of paper to the inspection
form for the mechanic or the cashier. Action: greater emphasis
in training and public awareness on handling the forms; instruc-
tion at troublesome stations by the CDR inspector.
• Inspecting mechanics or service writers attach other AIR Program
documents to the inspection form. Action: this problem virtually
disappeared when the full-volume phase began January 1, 1982.
Prior to that, the Department of Revenue PR 1390 form was required
and was frequently stapled to the inspection form.
39
-------
• Inspecting mechanics or service writers tear the form when
trying to separate the three pages at the perforations. Action:
no action yet, although thought has been given to removing the
. critical form narks from their proximity to the perforations.
This could be accomplished either by moving the perforated stub
to the right side or the bottom, or by moving the critical form
marks to the opposite side and end and keeping the stub at the
top.
Grease and dirt from the garage environment do not represent a signi-
ficant impediment to a high readability rate. An optical mark reader
can cope with small amounts of grease and dirt as long as they do not
appear in critical areas. A plastic template or overlay has been con-
sidered for use with the inspection forms. Its primary purpose would
be to guide the mechanic through the data blocks, rather than to keep
the form clean. Because few problems have been met which the template
could or needed to mitigate, and because of other priorities, the
template has not been developed.
Quality control checks exist to promote a high rate of readability.
When the Department of Revenue (CUR) inspectors conduct the regular
inspection station audits, they have opportunities to answer any
Questions the mechanics may have. Although it is not required, the
inspectors can spot check the forms at the stations to screen the
worst problems. A second quality control check occurs when the forms
are tallied, spot checked, and oriented for mechanical reading. The
forns are still in order by station so obvious and consistent errors
can be traced to the station. Problems are referred to Department of
40
-------
Revenue (CDR) inspectors for discussion with the licensed mechanics.
General cautions to stations and mechanics nay be made through the
training classes, the audits by the Department of Revenue (CDR), and
the technical bulletins mailed to the stations.
Readability has not proved to be a particularly problematic element
of the AIR Program data system. The Department of Health (CDH) Mobile
Sources staff estimates that about 95% of the inspection forms were
readable during the "blue phase." An estimate is necessary because
unreadable forms were rejected into the same bin as forms with data
errors, and it was difficult to distinguish between the two types.
Unreadable forms average 2.4% of the "green system" volume, ranging
from 1.0% to 6.5% on the weekly batches.
41
-------
4.0 DATA ANALYSIS
4.1 Purpose and Functional Elements
The ultimate purpose of collecting I/fl data is to learn how the
I/M program is functioning. A well-designed data handling component
should be able to monitor data and provide status reports.
The data analysis system developed for the Colorado AIR Program
is concerned with the data records read by the scanner and recorded on
the computer tape. Forms which were not physically scannable were
rejected at the scanner (section 3.4). That process - checking for
readability - is conducted by the scanner and for the purposes of this
report is not classified as data analysis. Readability is, however, a
consideration in determining a general "completion rate." Completion
rate could be defined as the quantity of data determined to be useable
for reporting purposes. Forms not readable detract from the completion
rate. Taped data records which contain errors also lower the completion
rate. Accuracy checking is the first function of data analysis and,
with readability, the remaining component required to determine the
program1s completion rate. Once a determination of data accuracy has
been made, data analysis can proceed to its concluding phase, that of
producing information in scheduled and special reports.
4.2 "Blue System" Accuracy Checking
With the July-December "blue system," the scanner edited the data,
discarding data from those forms which contained serious errors. Data
arriving at the computer were considered useable. The primary purpose
of the computer was to produce information. In July and August 1981,
approximately 30% of the incoming data was discarded by the scanner
43
-------
for readability and fatal data errors. Part way through the "blue
system" the criteria were relaxed, by which the scanner made its deci-
sion to accept or discard the data. Mechanics were issuing a certificate
of adjustment rather than compliance for vehicles passing after
adjustment. Because the emissions levels on forms with this type of
error were accurate, the scanner was instructed to accept the data
into the data base. Only about 10% of the incoming data was then
rejected for readability and fatal data errors.
4.3 "Green System" Accuracy Checking
Figure 3 depicts data handling under the current "green system."
A skeletal version of this system is now operating while the remainder
of it is beina developed. The first two processes (fatal error iden-
tification and questionable error identification) check the accuracy
of the data. In contrast to the "blue system," these edits are
completed by the computer, not the scanner.
The computer first reviews the record for each inspection. If the
scanner has indicated insurmountable problems with the data on a form,
the record is immediately rejected. Based on the fault chart in Table
3 developed by AIR Program staff, the computer assigns a value to each
record according to the number and type of data flaws it, finds. The
data elements assessed in this phase relate to information required to
set emission level standards (cutpoints). The assigned value - the
fatal error flay - is entered in one of the two fields reserved for
error flags for each record. Error messages relating to processing
are printed at. this stage if warranted (Figure 4). Using a cutpoint
of three and above, the current level of fatal errors appears to be
around 1%.
44
-------
Figure 3: DATA ANALYSIS
JANUARY 1, 1982 TO PRESENT
CDH
Data tape delivered to CDH Data
Services Section from scanner
to
03
01
Change format to
CDH compatible;
count forms
Error messages
Fara I error
Identification
(
CDH
Analyze Tor
information
*
1
I
Schedu led
spec i a 1
reports
^"
and
^ '
45
-------
Table 3: FATAL ERROR IDENTIFICATION
Make
X
(X)
(X)
(X)
(X)
(X)
(X)
(X)
(X)
Valu
Cylinders
(Number)
X
(X)
(X)
(X)
(X)
(X)
e(s) in
Model
Year
X
(X)
(X)
(X)
(X)
Error
HC
X
X
X
X
.a/
CO
X
X
X
X
Exhaust
Leak
X
X
Fatal
Error Flag
0
1
2
3
4
5 b/
6
7 y
8
8
9£/
g£/
JL/X indicates value is not available or is invalid or for "exhaust
leak" the value is 1; (X) indicates value may or may not be una-
vailable or invalid. No mark indicates value is available and valid
or for "exhaust leak" the value is 0.
Jl/HC or CO and at least one of number of cylinders or model year
unavailable or invalid.
£/HC and CO or exhaust leak and at least one of make, number of
cylinders, or model year are unavailable or invalid.
46
-------
Figure 4: PROCESSING SYSTEM ERROR MESSAGES
Message 1 SYNTAX ERROR
Message 2 SEQUENCE ERROR
Message 3 DATE ERROR
CURRENT DATE = MMDDYY
Message 4 MISSING OR INVALID RECORD COUNT
Message 5 AFTER READING zzzzz9 RECORDS FROM THE INPUT FILE AN
INVALID ID-SEQUENCE NUMBER WAS READ
Message 6 IMBALANCE ERROR
RECORDS-IN = zzzzz9
RECORDS-OUT = zzzzz9
REJECTED = zzzzz9
ENDING ID-SEQUENCE = 999999999
BEGINNING ID-SEQUENCE = 999999999
DIFFERENCE +1 = zzzzzzzz9
Message 7 IMBALANCE ERROR
RECORDS READ-IN = zzzzz9
RECORDS EXPECTED = zzzzz9
DIFFERENCE +zzzzz9
NB: A sequence of 9's (e.g. 999999) represents the location of a
numeric character string to be generated by the computer. One or
more z's preceding the 9's specifies suppression of leading zeros.
47
-------
Table 4: ERROR/PERFORMANCE REPORT - ERROR SECTION
ERROR DEFINITIONS - PRELIMINARY
NO ERROR FLAGS DETERMINED
MB: Footnotes appear on
last page of Table 4.
INSPECTION DATA-UNKNOWN
VIM
License Plate
Signature
Certificate No.
or
Inspection Cost
Test Date - month
day
year
INSPECTION DATA-ERROR
Inspecting Cost
Test Date (YYMMDD) ;.
VEHICLE DATA
Make
No. of Cyls.
Model Year
INSPECTION CHECKS
First Test - UNKNOWN
VI sual - yr
Enissions - P/F
Idle CO, HC
2500 CO, HC*
First Test - ERROR
Visual - yr
or
£
O
U_
*fc
ce E
— o
< +•
"" ~~
1
3
17
9
4
4
4
9
4
2
8
7
10
11
10
0)
.Q o
•
*
O
C£
(82+
(81-
0)
01
c
-
—
o
•a
o
2;
81+)
81-)
82+)
JC
c
O
4-
o
i/i
i/>
E
LU
•1- O
l/l O
i- ir\
— CN
U-
1
in
0
i/)
in
1
UJ
— 0)
ID —
C X)
•— ^_
b.
U)
8
VI
tn
1
UJ
— O
ID O
c in
— CN
L_
+-
c
I
I/I
D
•— >
•D
<
-*
- 0
a: o
^— <^
< CQ
•—
L.
-------
Table 4 continued:
ERROR/PERFORMANCE REPORT- ERROR SECTION1
ERROR DEFINITIONS - PRELIMINARY
•^v/ L.tn%vix • i_n\j-«s i**_ i L-IU IA ra^i/
MB: Footnotes appear on
last page of Table 4.
INSPECTION CHECKS (cont.)
Emissions - P/F
Idle CO, HC
2500 CO, HC*
Final Test - UNKNOWN
Visual - yr
or
or
Emissions - P/F
Idle CO, HC
2500 CO, HC*
Final Test - ERROR
Vi sual - yr
Emissions - P/F
Idle CO, HC
or
or
2500 CO, HC*
or
or
or
E
1_
O
u.
%
o: E
— L.
3 L.
OUI
^
•3
—
ID
>
2
O
ce
(P
(F
(P
(F
l
l
-
(82+
(81-
(P
(F
(P
(F
S,
c
•
Li.
Dl
1_
ID
0
>-
O
T3
O
s:
80-)
81 +
81+
81+
81+
81+
81-)
82+)
80-)
.c
§
-*-
ID
O
•—
M- -0
-g
1- V)
(U in
o —
>S1
s
s1
s
<^s
in
O
in
in
"i
LU
158
c ir>
— CM
u.
tds)
tds)
•t-
c
S^
4-
O
-------
NO ERROR FLAGS DETERMINED
Table 4: continued
ERROR/PERFORMANCE REPORT - ERROR SECTION
ERROR DEFINITIONS - PRELIMINARY
MB: Footnotes appear on
last page of Table 4.
INSPECTION CHECKS (cont.)
Adjustment/Repair -
UNKNOWN
Station Number
or
Mechanic Number
or
Adjustment Cost
or
Repair Cost*
Adjustment/Repair-ERROR
Station Number
Mechanic Number
Adjustment Cost
Repair Cost
or
INSPECTION RESULTS
First Test - UNKNOWN
Visual - P/F**
Emissions - Levels
Idle CO, HC
2500 CO, HC*
First Test - ERROR
Visual - P/F**
Emissions - Level s
Idle CO, HC
2500 CO, HC*
E
L.
o
u_
. «fc.
a: E
— L.
3 l-
OUJ
~°a>
W
D
<0
;>
%
0
ce
0
o
c
ID
K
*f-
O
-H
3
O
X
X
X
X
X
X
(D
r>
+-
3
s:
X
X
X
X
X
X
X
o
V
c
ro
CQ
(X
(X
(X
(X
(X
(X
(X
(X
X
(X
j£
c
(D
ca
c
o
z
(X
(X
(X
(X
(X
(X
(X
*+-
•f-
u_
O)
L
0)
>-
_
-------
Table 4: continued
ERROR/PERFORMANCE REPORT - ERROR SECTION
ERROR DEFINITIONS - PRELIMINARY
NO ERROR FLAGS DETERMINED
NB: Footnotes appear on
last page of Table 4.
INSPECTION RESULTS (cont.)
Final Test - UNKNOWN
Visual - P/F**
Emissions - Levels
Idle CO, HC
2500 CO, HC*
Final Test - ERROR
Visual - P/F**
or
Emissions - Levels
Idle CO, HC
2500 CO, HC*
or
CERT ISS'D DISAGREEMENT
Pass/ Adjustment
Other :
or :
or :
or
or :
or
or :
E
O
Lu
o: E
< -t-
14
15
14
15
.6/15
6/15
.6/15
.6/12
.6/14
.6/13
.6/12
6/12
a>
-O O)
ID ID
C —
O Li-
tn O
Q) u.
3 L.
OUJ
7
7
13,1
ID
^
ID
X
C
OL
4
O
82+)
81+)
81-)
80-)
81+
80-
.c
O
-H
<0
O
• •*—
•4- "D
.-
T3
^
S:8
«P. ^—
< CQ
blan
non 1
non
lank
ID
r~
•f-
o
.4
4
)
5)
.lank
6)
-------
Table 4: continued
ERROR/PERFORMANCE REPORT - ERROR SECTION
ERROR DEFINITIONS - PRELIMINARY
*Beginning July 1, 1982.
**Applies to each system (catalytic converter, fuel restrictor, AIR system) individually.
aWhere more than one error condition occurs on the same line, each condition applies independently
(non-exclusive OR).
error conditions are enclosed within parentheses all enclosed conditions apply simulta-
neously (AND).
cln cases where one or more error conditions is undetermined it will be assumed that there is no
error.
dRow Value - value of item to which row applies.
eBlank refers to item as a whole.
fTest includes both visual and emissions parts unless specified otherwise. P indicates passed
both visual and emissions parts, F indicates failed one or both parts.
9Model Year - 82+ indicates model years 32 and newer, 81- indicates model years 81 and older.
81+ indicates model years 81 and newer, 80- indicates model years 80 and older.
^Certification Issued - C indicates certification of compliance. A indicates certification of
adjustment. D indicates certification of denial.
^ 1 - Inspection cost > $10.
2 - Test date >^ current date.
3 - Adjustment cost > $15.
4 - CO levels > 15%.
5 - Repair cost < $100 or blank.
6 - Repair cost > $100.
-------
Next the computer reviews each record for questionable errors.
The criteria with which the computer is to judge the seriousness of
the questionable errors is under development by the AIR Program staff,
but a preliminary version is provided in Table 4. The error flag
value for each type of error has not been established yet. Basically
this system assigns a value of 0-99 to each record. A minor error
will be assigned a low number; a significant error, a higher number.
Combinations of errors will be assigned still higher numbers. The
severity of each error is based on the judgement of the AIR Program
staff according to the Program's established data analysis needs. The
questionable error flag is entered in the second error flag field for
each record. Each record now contains the data which were read from
the inspection form by the optical mark reader and two error flags.
The first, fatal errors, is for data critical to setting emissions
standards (cutpoints); the second is for data with less serious errors
(questionable errors), but errors which are indicative of the general
reliability of the data on an inspection form.
The two error flags will be used in the following manner.
Depending on the purpose of an analysis, two "reliability levels" may
be selected. The questionable error flag indicates basically how
trustwprthy the data on that form are judged to be. The fatal error
flag indicates how valid the critical data elements are. A count of
the number of inspections on the data tape would include all forms
(questionable error flag of 99 or lower, fatal error flag of nine or
lower), but an analysis of adjusted emissions by model year might
require data rated at 50 or lower, with a fatal error flag of two or
lower. Make and model distributions of emissions levels might include
53
-------
a questionable error flag of 10 or lower and a fatal error flag of
three or lower. The flags nay he altered independently of each other.
When operational, this system will provide an indication of how
accurate the captured data are. It will allow subdivisions of the
data bank based on purpose and degree of accuracy required. It will
allow much greater analysis of the accuracy of the data, and thereby
enhance the credibility of the analyses produced.
To the knowledge of the Colorado Department of Health (COM) Data
Services Section, the error identification system being developed by
the AIR Program staff is unique. It is difficult to develop and. will
require extensive testing. The time and effort involved in developing
such a reliability rating system must be balanced against the value of
the information gained and the needs of the program. For the AIR
Program it promises significantly greater flexibility in handling the
data captured hy the optical mark reader.
One of the products of the error identification procedures is a
series of error reports developed after the second edit. These sum-
marize the types of errors found in the data and organize the infor-
mation by inspection station and mechanic. Processing error messages
(Figure 4) are produced when the records are first checked following
receipt from the scanner, but the reports generated after the second
edit are the first in-depth administrative reports furnished by the
system. Figure 5 is a nock-up of a batch error/performance report
under development. The report consists of two sections. The first,
the error section, reports the types and quantities of errors the
mechanics are making when they complete the inspection forms.
Mechanics or stations with high error rates will be marked with
54
-------
BATCH ID RANGE
999999999 - 999999999
Fi gure 5
COLORADO DEPARTMENT OF HEALTH
ERROR / PERFORMANCE REPORT
ERROR SECTION
MMM 29, 1999
PAGE 999
STATI ON
NUMBER
MECH
NO.
0001
ITEM 1,3,4,9, 2,
FORMS NOS: 17,CERT NO. 7,8
READ FORM INSP DATA VEH
ERRORS UNK ERR DATA
(ALL ERROR RATES EXPRESSED AS PERCENT)
INSPECTION CHECKS ----- - -INSPECTION RESULTS- -
10YR,11PF
FIRST TEST
UMK ERR
14YR.15PF
FINAL TEST
UNK ERR
12,13
ADJ/REPAIR
UNK ERR
10PF,11#S
FIRST TEST
UNK ERR
14PF,15#S
FINAL TEST
UNK ERR
CERT ISSUED
DISAGREEMENT
16
PASS
/ADJ OTHR
999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
1000
9999 999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
1000 ALL 99999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
1001
0001 999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
z z z z z z zzzzzzzzzzzz
1999 ALL 99999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
1XXX STATUS 999999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
2000
Z0001 999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
z z z zzz zzzzzz zz zz z z
9998 ALL 99999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
9XXX STATUS 999999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
9999 ?STA#? 99999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
ALL STATUS 9999999 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9
-------
Figure 5
01
BATCH ID RANGE
999999999 - 999999999
NUMBER
MECH
NO.
1000
0001
• y
9999
1 000 ALL
1001
7 0001
/ X'
1 999 ALL
1XXX STATUS
2000
7 0001
FAIL
RATE
RET
ALL
99.9
./
99.9
99.9
99.9
/-'
99.9
99.9
99.9
# OF
VEHS
RTST
ALL
9999
X'
9999
9999
9999
^x
9999
9999
9999
C 0 L 0 R A
DO DEPARTMENT OF HEALTH
ERROR / PERFORMANCE REPORT
PERFORMANCE SECTION
AVG EMISSIONS EM REDUCTION
FIRST FINAL AVG %
RTST
ALL
09.9
X
09.9
09.9
09.9
X
09.9
09.9
09.9
RTST
09.9
X
09.9
09.9
09.9
X
09.9
09.9
09.9
RTST
ALL
09.9
x
09.9
09.9
09.9
09.9
09.9
09.9
RTST
ALL
99.9
X
99.9
99.9
99.9
X
99.9
99.9
99.9
# OF AVG EMISSIONS EM REDUCTION
VEHS FIRST FINAL AVG %
RTST RTST RTST RTST RTST
ALL ALL ALL ALL
9999 09.9 09.9 09.9 99.9
X X X X X
9999 09.9 09.9 09.9 99.9
9999 09.9 09.9 09.9 99.9
9999 09.9 09.9 09.9 99.9
x x^ / / /
9999 09.9 09.9 09.9 99.9
9999 09.9 09.9 09.9 99.9
9999 09.9 09.9 09.9 99.9
AWT 1 If* -API IMMTC
INSP ?'OF AVG
COST VEHS COST
RTST
ALL ALL
99.9 99.9
X X
99.9 99.9
99.9 99.9
99.9 99.9
Z_ , .-y
X
99.9 99.9
99.9 99.9
99.9 99.9
RTST
ALL
99.99
x^
99.99
99.99
99.99
^x"^
99.99
99.99
99.99
MMM 29, 1999
PAGE 999
RFH PPPA 1 PC
% 81+ AVG
VEHS COST
RTST
ALL
99.9
z.
99.9
99.9
99.9
X
99.9
99.9
99.9
81+ R
81+ A
999.99
^^
999.99
999.99
999.99
^^
999.99
999.99
999.99
9998 ALL 99.9 9999 09.9 09.9 09.9 99.9 9999 09.9 09.9 09.9 99.9 99.9 99.9 99.99 99.9 999.99
9XXX STATUS 99.9 999999 09.9 09.9 09.9 99.9 999999 09.9 09.9 09.9 99.9 99.9 99.9 99.99 99.9 999.99
ALL STATUS 99.9 9999999 09.9 09.9 09.9 99.9 9999999 09.9 09.9 09.9 99.9 99.9 99.9 99.99 99.9 999.99
-------
asterisks for remedial assistance from the Department of Revenue (CDR)
inspectors. The second section, performance, concentrates on consumer
protection aspects of the inspections and adjustments. By comparing
the performance of a mechanic or a station with the average performance
("ALL STATUS"), the quality of adjustment and repair can be monitored.
Again, asterisks will flag abnormal rates. Copies of the batch error/
performance report will be distributed to Department of Revenue (CDR)
inspectors, who will conduct field visits at those stations indicated
on the report as having aberrant error or performance rates.
The last stage of accuracy checking in the "green system" occurs
when the batch data are merged with the master, or entire extant, data
bank. The master tape used is the computer tape generated when the
immediately previous data batch was merged; it is one week old. Two
back-up copies of the master tape, two and three weeks older than the
master, are retained. The three week old tape is rotated to a separate
storage vault each time a new master tape is created. At the nost,
three weeks' worth of inspection forms would have to be rescanned to
recapture data in case of a catastrophe. When the batch tape is merged
with the master, the program counts the number of records, puts them in
order by the sequential scanner number, and checks for duplicate
sequential numbers. This phase produces error condition reports.
Figure 6 is a mock-up of a graphic description of the quantities and
types of errors found in both the batch and new master data files. It
reports the number of records found unreadable ("bad scans"), then
depicts each type of error as a section of the "pie." For quick
reference purposes graphic depictions are very useful.
The accuracy checking accomplished by the "green system" identifies
abnormalities based on statistical assurances. The abnormalities can
57
-------
REPORT DATE
XXX 29, 1999
Figure 6
COLORADO DEPARTMENT OF H
INSPECTION/MAINTENANCE PROGRAM
WEEKLY BATCH SUMMARY
E A L T H
en
00
BATCH TOTALS
BEG INNI NG ID 999999999
END I NG ID 999999999
RECORDS IN zzzzzz9
BAD SCANS zzzzzz9
RECORDS OUT zzzzzz9
FORMS WITH NO ERRORS zzzzzz9a
FORMS WITH QUESTIONABLE ERRORS
AND NO FATAL ERRORS zzzzzz9
FORMS WITH FATAL ERRORS...zzzzzz9
****#******»*»*
»**»#====== •«••*****
***»++========i"!!.'!... .****
*»*+++++======== ***
**+++++++======== **
**-H-+-M-++++======== **
**+-H~H-++++++======= **
**+++++++++++++====== +++**
**+++ f+++++++++++===== ++++++**
FORMS WITH NO ERRORS :++++
FORMS WITH QUESTIONABLE
ERRORS & NO FATAL ERRORS:....
FORMS WITH FATAL ERRORS :====
NEW MASTER TOTALS
OLD TOTAL RECORDS
NEW TOTAL RECORDS zzzzzz9
FORMS WITH NO ERRORS zzzzzz9
FORMS WITH QUESTIONABLE ERRORS
AND NO FATAL ERRORS zzzzzz9
FORMS WITH FATAL ERRORS...zzzzzz9
*»*****»****»»»
»*»**======> **»»»
»»**++========,]]*.*!....»*»*
***4-++++=s==ss: = = #*#
**+++++++=== = == = = **
**++++•*•.++++======= = **
*»+++++++++•(•+= = = ====. **
**++++++++++•(•++====== -H-+**
**++-H-H"H-f+++-M-+===== +-M-+++**
**+++++++++++++++++====
*+++++++++++++++++++===
*++++++++*+++++++++++==
+++++++++++**
++++-H-+-M-+-H-++*
+-H-+++++++-M-++++*
*++++++-(-++-l-++++++++++-H-= .
+++-M- •*-•(•+•»•++++++++++*
*+++++++++++++++++++++++++++++++++++++++++++++*
**++++-(• -H-I-+++++-I-++-H-+++-H-+++-I-+++++++++++-H-+++**
*+++++++++++++++ +++-M-++-H-I-H-+++++++-H-+++-I-+++*
**++++++++•++•)•++++++ ++++++ ++-I-+++ +•»•++++**
l-4- +++***
*****+++++++++++-)•+*****
* **#*»*«-»*#****
»*»****»*»***»»
FORMS WITH NO ERRORS zz9*
FORMS WITH QUESTIONABLE ERRORS
AND NO FATAL ERRORS zz9?
FORMS WITH FATAL ERRORS zz9$
FORMS WITH NO ERRORS zz9?
FORMS WITH QUESTIONABLE ERRORS
AND NO FATAL ERRORS zz9$
FORMS WITH FATAL ERRORS zz9$
a"Forms with no errors" may be either error free or with errors assigned values lower than the outpoints specified for this report.
-------
rest either with the inspection station or mechanic, in which case it
becomes the focus of attention for enforcement, or with the data
handling procedures, in which case Mobile Sources and Data Services
must correct the procedures. Only extensive testing will prove the
full capabilities and reliability of the accuracy checking element of
the full volume data analysis system.
4.4 "Blue System" Information Producing
Data analysis for the "blue system" was devoted almost exclusively
to providing information. One report, the weekly "AIR Error Report,"
was custom programmed in COBOL, with the other programs relying on the
capabilities of SPSS. The brevity of the "blue system" (six months)
precluded extensive custom programming. Appendix G includes a summary
of "blue system" reports, copies of two reports, and report schedules.
Several bar graphs are also included; these were produced by trans-
ferring computer generated summary data to a Hewlett-Packard 85
graphics terminal not operated by the Data Services Section.
4.5 "Green System" Information Producing
With the "green systen" three reports will be supplied on a
routine basis:
monthly station performance
bimonthly model year performance
quarterly distributions and test results
In most cases the type of information provided regulates the frequen-
cies at which the reports are produced. Model year performance can
vary so much, for example, that bimonthly reporting compensates for
short-term variances. While the complete data analysis for the
"green system" is being developed, interim reports are provided on a
59
-------
frequent basis using a preprogrammed software package. Report topics
Include:
1. emission reduction,
2. pass/fall rates on the entire data base,
3. retest rates,
4. the number of vehicles falling 1n the first and second test.
All of the programming mentioned so far, except that for the
Interim reports, has been custom written for the needs of the AIR
Program. Developmental time, as shown in Figure 7, can be quite
lengthy, and unexpected delays can easily double the amount of time
required. Using preprogrammed software can decrease the burden placed
on the programming staff. The AIR Program relies on one of the
available software packages, the Statistical Package for the Social
Sciences (SPSS),2 for almost all requests for special reports. SPSS
is a package designed for social science data but is useful 1n most
fields. It is excellent for data manipulation, provides for flexible
labeling and saving of data files, and is especially valuable when
many different analyses are to be run on the same data. Its capabili-
ties include descriptive statistics, contingency tables, regression
analysis, partial correlation, analysis of variance, discriminant ana-
lysis, and factor analysis. The AIR Program staff finds the SPSS
package somewhat less efficient than other packages when used to per-
form significant numbers of calculations or data transformations.
Costs for programs like SPSS vary greatly; $3,000 originally and
2SPSS is a trademark of SPSS, Inc. of Chicago, Illinois, for its
proprietary computer software. No materials describing such software
may be produced or distributed without the written permission of SPSS, Inc.
60
-------
Figure 7
DATA ANALYSIS PROJECTIONS
Date Prepared: February 2, 1982
Week Ending: 1/8 1/15 1/22 1/29 2/5 2/12 2/19 2/26 3/5 3/12 3/19 3/26 4/2 4/9 4/16 4/23 4/30
BATCH DATA
Scanner
Program
Output: Tape
Printout - Form Counts
Error Level 3
Program
Output: Tape
(Printout - Error Messages)
Error Levels 1 and 2
Program
Output: Tape
Level 3 Error Summary Report
"Weekly" Error/Performance Report
Merge
Program
Output: Tape
Batch and New Master Summary
(Error Condition Report)
CUMULATIVE DATA
Routine Reports
Monthly Station Performance - Report
- Program
Bimonthly Model Year Performance - Report
- Program
Quarterly Distributions and Test Results - Report
- Program
Special Reports
(To be determl ned)
.
a
a
a..
a
a
a
a..
a.
a
a
a
b
crv
Concept . . . Design Development .. Testing
aAwa1ting approval by Mobile Sources Section
bTo be revised when Department of Revenue adds mechanic numbers to tape.
Operational
-------
$1,000 per year to use it 1s typical. A more complete description
of SPSS appears 1n Appendix H. It should be noted that SPSS does
require a certain level of computer and statistical knowledge and
experience. A knowledgeable staff member will be able to manipulate
the SPSS program to extract needed information. A recent AIR Program
analysis, for example, compared emissions reductions of models with
sealed and unsealed idle mixture screws. The analysis was accom-
plished with the SPSS program by carefully excluding makes and model
years from the pertinent data base, rather than by custom designing a
program over two to four weeks or more. Other analyses which can be
accomplished include:
• if all inspected vehicles had to pass for carbon monoxide, what
would reductions be,
• what reductions would be obtained in a carbon monoxide only
program with the same standards as now, and
• how large a reduction is obtained on vehicles still falling after
adjustments.
4.6 Staffing, Equipment, Costs, and Other Computer Systems Considerations
Staffing for the data analysis component of the AIR Program comes both
from CDH Mobile Sources (three Full Time Equivalents) and CDH Data Services
(one FTE). One Mobile Sources staff member is responsible for data
functions. He is a programmer and additionally is familiar with the I/M
program in general so he can, for example, design the fault chart for the
questionable errors. One other member coordinates the data collection
and a third is the data clerk who checks and orients the inspection forms
before scanning. Three individuals in Data Services have contributed to
aspects of the AIR Program. With their one FTE for the AIR Program
they provide expertise in the areas of the NCS scanner, COBOL, SPSS,
62
-------
statistics, and general project supervision. As the AIR Program
completes developmental activities, the four FTE's should decrease to
two FTE's. Programming will decrease to a minimum, requiring only a
part of a Mobile Sources FTE and no Data Services FTE. Although the
data clerk will be busy, .5 FTE of the data coordination position should
be available for other duties.
The Colorado AIR Program uses a Data 100 (Northern Telecom) system
as a remote job entry (RJE) station. This system accepts jobs for
the central processing unit and prints reports. It should be noted
that the standard report graphics are being printed on a non-graphics
terminal. This complicates programming tremendously for pie charts.
Some bar graph programming had been developed for another project and
is being applied to this project. An IBM 3033 functions as the
mainframe. Actual analysis occurs here, although the IBM 3033 is not
known as a system especially adept at numbers analysis. As the data
files increase in size, processing time may become lengthy. Some
general concerns for computer systems needed to handle projects like
the AIR Program are as follows:
• Can the system handle large files?
• Is the operating speed acceptable for the types of analyses to
be accomplished and the types and volumes of data to be analyzed?
• Can the system support SPSS or one of the other statistical
packages available?
• What types of storage are available on the system?
The AIR Program data system is handled in weekly batches for data
update. All data are stored on tape and are batched, rather than being
stored on-line on disk. This greatly reduces the costs of the data
system. January 1982 is not a representative month for data costs
63
-------
because full volume analysis had not begun and the data base was
small. It does serve, however, to indicate the savings afforded by a
batch system. Costs, excluding SPSS costs, totaled $163. Central
processing unit (CPU), or actual analysis, time was only $20 at a rate
of $263 per hour. $93 of the $163 was for program development; part
of the remaining charges was for hanging 22 tapes during the month.
On-line charges for a similar amount of effort could have been six to
eight times as much, and storage costs for the 80,000 to 100,000
records on file in January could have been considerable. As a data
system develops, costs are expected to increase with the volume of data
on file and the increased amount of processing time required to ana-
lyze the data.
4.7 Administrative Detail
All data analysis requests, whether from the Department of Revenue,
the Legislature, or reporters, are handled by the Department of Health.
Routine reports are typically for purposes of enforcement or program
administration (setting standards, reviewing reductions, etc.) and
are provided to agencies and groups as needed. Included are the Air
Quality Control Commission, Department of Revenue, Legislature and
legislative committees, EPA, County Health Officers, and media repre-
sentatives.
The only serious problem encountered in the data analysis com-
ponent of the AIR Program has been that of scheduling. It has proved
extremely difficult to accurately estimate how long to allow for deve-
loping programming. Because Mobile Sources and Data Services work
together closely, coordination of their efforts to eliminate waits on
64
-------
each other has seemed to present the worst problem. Problems associated
with equipment failure, Illnesses and vacations, and programming
difficulties are to be expected. Extra time needs to be built into
planning schedules to accommodate for them. If contracted equipment is
being used, extra time should be allowed for data staff to become
familiar with it.
65
-------
5.0 SUMMARY
5.1 Description
The Colorado Automobile Inspection and Readjustment (AIR) Program
uses machine readable forms to collect emissions inspection and
readjustment data. This makes it unique among inspection/maintenance
(I/M) programs currently operating. Reliability and cost of the opti-
cal mark reading system were the primary reasons for using it rather
than other automated-systems. Over 3,000 licensed mechanics have been
trained to complete the hand-coded inspection form. The forms are
collected during monthly station audits by Colorado Department of
Revenue (COR) inspectors. The services of a NCS Sentry 7010 optical
mark reader operated by Denver Public Schools have been contracted to
scan the forms, converting the data fron handwritten narks to a com-
puter tape. Data on that tape are analyzed by a state operated com-
puter, an IBh 3033. Reports are available at regular intervals to
satisfy the two primary questions regarding the AIR Program:
• are consumers being adequately protected, and
• are the desired emissions results being achieved?
Special reports are available upon request and are produced with a
pre-programmed statistical package, the Statistical Package for the
Social Sciences (SPSS).
The I/M program in Colorado has been implemented in three phases.
The earliest consisted of data gathering from studies, conducting a
limited number of voluntary inspections, and developing a mechanic
training program. The AIR Program cane into being with Senate Bill 52
and Became operational July 1, 1981 (phase 2). This six month phase
67
-------
included aproxinately 140,000 vehicles which were fleet operated, new
registrations, or vehicles undergoing change of ownership. The
July-December 1981 program is referred to as the "blue system," from
the blue inspection form. The "blue systein" afforded the AIR Program
staff opportunities to test, procedures and regulations before the
full-volume I/M program began January 1, 1982.
The third phase of the program, called the "green system" because
of the green inspection form, is now partially operational. About 1.4
million inspections will be conducted annually at a projected volume
of 20,000 inspections per week. The quantities of forms received and
scanned have varied fron a rate of around 3,000 per week in January, to
11,000-12,000 in early March, to over 20,000 form per week in mid-April
Data on the resulting data tape undergo computer analysis, but the full
report schedule is not yet operational. A sophisticated error iden-
tification system is being developed which will nake the captured'data
more useahle and also provide a "reliability factor" to the analyses.
This error identification is the ^ost significant change between the
"blue system" and the "green system." Both of the systems have been
surprisingly successful in capturing data, as well as in capturing
seemingly accurate data. Vlhen the "green system" is fully opera-
tional it will be able to report just how accurate the data records
actually are.
5.2 Critical Aspects of the AIR Program
The most serious problem experienced by the AIR Program appears to
be that of projecting the developmental time needed by various com-
ponents of the progran, especially data analysis. Because it is an
innovative program, it is difficult to foresee where more time should
68
-------
be built into the projections. Involving several groups (CDH, CDR,
Air Quality Control Commission) in aspects of the AIR Program means
that no single resource can be assigned to all activities. Each
activity must be coordinated so that it is accomplished as its priority
requires and as the several resources are available. All of the
resources need to be flexible. Program changes can - and should - be
initiated internally as problems are discovered and solutions are
found. Changes can also cone from outside sources in the form of
legislative action. A viable I/H program can respond to specific
changes with a m'nimun of disturbance to the rest of the program.
Training the mechanics to complete the coded inspection form has proved
valuable, whether one-on-one as was done in the earliest stages of the
AIR Program by CDR inspectors or in classes as has been done for
roughly two-thirds of the program. Training appears to have contri-
buted significantly to the high readability rate of the forms.
One measure of the success of a data handling system can be
strictly numerical: how much information is gathered and useable?
The "green system" scans data from more than 20,000 hand-coded inspec-
tion forms each week. On the average, only 2.4% of the forms are lost
because of readability problems. Of the remaining 97.6% of the data,
only 7% contains fatal errors when the fatal error cutpoint is
established at three or above. Neither this 7% nor any percent of the
'data -identified as containing questionable errors (when that system is
operational) will be eliminated from the data base. The full 97.6%
volume is available for analysis by sliding the fatal and questionable
error cutpoints on their respective scales according to the types of
69
-------
analyses needed. A large volume of inspection data is collected and
an unusually high percentage of that data is successfully captured
for data analysis.
Another measure of success relates to what is done with the useable
data: does data analysis meet the reporting needs of the program?
Again the AIR Program seems successful. The inspection forms are
approximately 30 days old when received for processing. Ten days are
used for processing each batch completely. Data results are therefore
no more than two months old. Custom programming produces in-depth
reports for each hatch and on a routine schedule. Usino a pre-
programmed pacl'.age such as the Statistical Package for the Social
Sciences (SPSS) allows a great deal of flexibility for special
reporting.
The Colorado Alfl Program has been, and continues to be, developed
through a "learn as you go" process. This seems to be the key to its
vitality today. The data system efficiently collects emissions data
and processes those data to provide up-to-date information for a
variety of needs. The primary purpose of an I/H data system is to
answer the question, "How well is tin's I/f! program operating?" By its
own example the Colorado AIR Program data system gives credence to the
answers it provides.
70
-------
APPENDIX A
73
-------
The following text is from the most recent (August 7, 1981) draft
of the Arthur Young & Company information system report for the
Colorado Air Pollution Control Division. Only those sections
pertaining to the I/M program subsystem have been reproduced here.
75
-------
2.5 MANAGEMENT INSPECTION/MAINTENANCE PROGRAM
2.5.1 Introduction
On July 1, 1981 Colorado's vehicle inspection and
maintenance (I/M) program will begin on a change-of-ownership
basis. Mandatory inspections for all affected vehicles in the nine-
county I/M program area is tentatively scheduled to begin on
January 1, 1982. Relevant details of the program are described
below:
• All 1968 and later light duty vehicles in the program
area must receive a certificate indicating compliance
or adjustment with the I/M program as a prerequisite to
the current safety inspection.
• The emissions inspection will be conducted in licensed
private garages. Vehicles with emissions exceeding the
standards or "cut points" established for a particular
model year must be adjusted by a licensed mechanic
according to manufacturer's specification in order to
receive a certificate (or if the adjustments are
conducted by someone else the vehicle must "pass" on a
retest).
• Regardless of who makes the repairs, a second test is
performed on all vehicles which are repaired. The
purpose of the retest is to indicate the change in
carbon monoxide and hydrocarbon levels over the first
test.
• The program is designed to fail no more than 40 percent
of the model year 1968 through 1974 vehicles and 30
percent of the 1975 and later vehicles.
• The program will be administered by the Mobile Sources
Section of the Colorado APCD in conjunction with the
Colorado Department of Revenue (which currently
administers the annual vehicle safety inspection
program).
• APCD is responsible for training Department of Revenue
personnel to audit inspection and repair facilities, to
train inspectors and mechanics, and to provide program
surveillance. APCD also must continually reevaluate
cut points to avoid exceeding the 40 and 30 percent
failure rates set as a maximum by the state law
(Colorado Revised Statutes 42-4-306-5 et seq^) which
77
-1- June 8, 1981
-------
authorizes the program. This law also requires the Air
Quality Control Commission to report annually the costs
and effectiveness of the I/M program.
These requirements impose a variety of data gathering obligations
upon the Mobil Sources Section. Most of these can be handled
through a Data Management Program. Exhibit 2.5-1 depicts the data
flow which is described below in greater detail.
2.5.2 Information Source
The source of the data is a one-page form which is filled
out by the mechanic for each vehicle tested. Exhibit 2.5-2 is the
form which is currently being considered. This form will be read by
an optical scanner.
2.5.3 Data Collection Activities
The data sheets will be filled out by hundreds of licensed
mechanics in the field. The Department of Revenue will supply the
mechanics with the forms and will collect the forms on a monthly (or
more frequent) basis. Approximately 1.5 million vehicles will be
inspected each year.
Typically, the mechanic will begin by completing the items
which identify and describe the vehicle (license number, model,
engine size) and identify the inspection station, mechanic, date,
etc. Both a visual check (an inspection to determine the presence
of pollution control equipment and exhaust system integrity), and a
tailpipe emission inspection are conducted. The CO and HC emissions
are then entered onto the form. If the vehicle passes both
inspections, the mechanic checks off the compliance box on the form.
Approximately one-third of the vehicles are anticipated to
fail the inspection, whereupon adjustments (and in some cases,
repairs) are made and tailpipe emissions retests are conducted.
These results and the cost data related to the repairs and
adjustments are entered onto the forms.
78
-2- June 8, 1981
-------
UD
DATA INPUTS
Inspector
Prepares
Data Form
Input Data Categories
a) Vehicle Identification
b) Vehicle Characteristics
c) Administrative
d) Cost
e) Inspection
f) Repair
DATA PROCESSING
Department of Health
Data Processing Stores
and Processes Data
from 1.5 Million Forms
Each Year
DATA OUTPUTS
Monthly Reports
Quarterly Reports
Annual Reports
EXHIBIT 2.5-1. I/M PROGRAM MANAGEMENT
DATA FLOW DIAGRAM
-------
EXHIBIT 2.5-2 PROPOSED DATA FORM
t OM>< eigmta 1:1
-rr». i».m
COLORADO DEPARTMENT OF HEALTH
"Air PROGRAM INSPECTISN/ntADJUSTMENT
REPORT
VIHICLE iot;;iif,CATION
cw C\
,t-n* O!
,'ci.iv O
.Oil O
0 LICENSE HATE
1 1 1 1
GGGGOO
GOSO0O
GOOOOG
IGGG0GG
©GG®GG
oooooo!
®®G®GG
0 DATE Of TEST
T
[ 00
I0000000000
1
1 • 0000
I000000000©
1 ©
I000000000©
0STATIOM
NUMBER
1 1 1
0©00©0©00®
000©©00©0©
©00©00©©0©
0©00©0©0.0©
©
• ItriMC 1
kgulll
1 1
0©000©000©
000000000©
©00000000©
(i)00C-)00000®
O " "
1
©®
OG
GG
GG
GG
GG
G®
GO
G®
GG
"©I
©00&0000©©|
©0 1
1
(•} Clll1l'ICMlfl:< ISSUED
CUMTl.iiKC£ O
AOJUJTME'J.' G
OtNifO G
O RETES? C.-.;:::c:;S ifvu
*« CO t !>•• "C
i '
GG.G
00.0
G.G
! l 1
|fl.« C! ®G®©®0
• _ *•• I I >•* *** y^ ^^. *"^ ^^
C!
MC
ntos O
ITU
rt»M
GOOOOG
'©0©©0©
GOGGGG:
,530.30®
JG©®©G®
'©00GO©
©G0©©©
0OCOOO
|s^C 1000GS©
!S-JCA O jGGGGGGj
i'ov C S0G<££3
Gi GGGGG©
G! GGOGGG
Oi 1000000
©.©I'" :"
.... _®;
©GO!
G©G!
6
IN
COSTS UN COLLARS!
:r
©®
GG
G
©
S
©
0
©
©
1
A3J
0®
GO
©
©
G
©
S
G
©
DttUiMa M'AIII
lAllOU
0000000000
00O000000©
000000000©
FARTS
00000000®©
©000000000
00©00©000©
© RETEST
VISUAL n:Mt;iDn
FAIt|0
CATAHTICCar.VERIjH G
AIR SVSUM O
EXHACST
JVSTEV INTEGRITY Q
©Gii!», GGG
G.®!""" j ©GG
G.O '• GOO
©.G' i GGGj
© P*SS G
© fAlt 0
O I certify Ihil 1 havt (eifotncd thit m;r;etion
and any required idjwitmiriti in acciv'ir:: «'J.
the iulc> and guldtlinci u( tin Colctc'o AIR
Program
SISKATURE Of LICEMStO EV.:SSIO.% MECHV-.C
©
, IM.SSI
VIKIZLl
TIAIi
UiJ-71
1)11.14
l!»->l
Illl. 11
>«>» t
:n; STANOAROS
CO
>o\
tot
ssv
it\
in
_at_
U00),«
UOOi,.
1031,.
taou*
iS7iM INTEGRITY Q
fmST TEST (MISSIONS KVELS
xcg
®©.©
GGG
GG
GO
GG
G.G
G-G
GG
Ittll.
!• Ik*
U( me III
©
O
••«
©©©©
GGGG
GGGG
GGG
GGG
GGG
GG®
GGG
©G©
VOIUMAH.T «:n;s
ADJUST
ME:r -.-Mr ~~\
TUf.'E-l?
CARCURET0.1
AIR CLEANER
CHOKE
OTHER
O
O
O
O
O
'ASS @
»All 0
HOME ADJUST
•\CO
®®G©
GGGG
GGGG
GGGG
GGGG
GO®®
GGG©
GO©©
©GGG
•SSGG
CDC.!.©1
GG-TG
GOG®
©SGU
©OS®
©GGG
USE A NUMBER 2 PENCIL ONLY ON THIS FORM
FOLLOW INSTRUCTIONS ON REVERSE SIDE
THIS COPY TO DEPARTMENT OF HEALTH
PACE 1
80
-------
EXHIBIT 2.5-2 (Page two)
INSTRUCTIONS TO LICEKCE-J Erji'.SSiOi'iS P.IECMAHISS
(ONLY a Licensed Emissions Mechanic may perform ti:is inspection.)
NOTICE
The first copy (top) of this form will be scanned electronically
Please keep the top copy free of dirt or grease.
Do NOT fold, staple, spindle, or mutilate lop copy.
Do MOT make ANY marks on the top copy other than specified in directions.
You may make notes on the second or third copies ONLY.
Use ONLY a NUMBER TWO pencil on this form.
Whcrj boxes are provided put one and only one number or letter in each box. if there are more tnxcs
than numbers or letters put enough zeroes in Iron! (to the jeft) to fill up the extra buxes. For a«c:.i;n by Die Fcr;:.i
who does the work and relumed within 10 days for a frei reinspcctitn.
BEINSPECTIOM
E If the vehicle was adjusted or repaired by a licensed Emissions Mechanic, ccpy the s'atirn and rcchar.ic': iiceist •••ir.;-
beis Iro.n the "A4|ustment Venlicat.gn" lorn under rum O. attach one ccpy ol the "Aj;u;..iieni Vi'ilicmn lo'rr. to \.-.~,t
1. and 30 on to Step G.
F. If ihe vehicle was adjusted or repaired by a non-licensed person, complete item 0. II the vehicle passes y> on 13 ,"•:»
G. II il fails, make Ihe adjustments or repairs required and go on to S:«p G or mark "Oemea" O. sign the lorm J'.
punch CO ©. and 'Date of Test" on pages 2 and 3. and give both halves ol page 3 la the vehicle o.vner
G if ihe vehicle failed the first test visual, do the rttesl visual inspection 0. il n passed ihe first test visual go on to s:c; H
H Oo the retest emissions inspection G. (mark PASS or FAIl for CO and HC].
I. Record trie amount charged lor the inspection, thi adjustments (if performed), and Ihe labor and parts casts lor rt;i:i^'1
repairs ©
J. Record any voluntary repairs O.
K. II th« vehicle passes both the visual and the emissions retests. mark "Compliance" O. sign the form G. tunch
G>. and "Data ol Test* on pages 2 and 3. and give both halves ol page 3 to the vehicle ov:ner.
L II the vehicle passes the visual relest but fails the emiiiioni rtiest (and lor 133 fs and newer. $10033 ais sp:nt :n
imimoni repairs! maik "Adjustment" 0. sign the form ®. punch | CJA j Q. and "Date ol T-sl" un pgrj,: ; jUd
}. and give both halves ol page 3 lo the vehicle owner.
M II KI Chicle fails me visual retest tor. for I98t's and newer, fails ihesrnss.cns rtll!1, .nt |,iS ,hln S;00 M.. , ,
on emissions repairs) mark "Denier 0. sign the lorm ©. punch [ CD | ®. and 'Oa:e ol Test" on pj;s; 2 :i(!
3. and give both halves ol page 3 lo the vehicle owner.
81
-------
The volume of data to be processed will be large. If each
of the 1.5 million vehicles is inspected there will be 1.5 million
hard copy forms collected annually. However, not all data elements
will be completed on each form because vehicles which "pass" the
visual and emissions test are not required to be adjusted and
retested. Only about one-third of all vehicles are expected to
fail, thus two-thirds (or about 1 million forms) will be partially
completed. The type of adjustment or repair and where the repair or
adjustment is conducted will determine whether or not other elements
are entered on the form.
In Section 2.5.8 (data categorization) estimates of the
volume of data to be processed for each data element are given.
Based on these estimates, a total of 31.3 million elements will be
processed the first year of the program increasing at a rate of
about 8 percent per year.
2.5.4 Retention of Data
Completed forms will be delivered to the Department of
Health data processing section where they will be read by an optical
scanner and the data will be stored on tape. The forms will be
stored in a warehouse for at least two years. The tapes will be
retained in active storage for one year and retained indefinitely
thereafter.
2.5.5 Data Calculation/Processing
The inspection data will be used to produce monthly,
quarterly, and annual reports which present the number of failures
and emissions reductions by vehicle type, repair station/mechanic,
and the program as a whole. These data are processed by the data
processing station in the Department of Health. Specific data
manipulations are described below:
a) Counts, percentages, ratios, and emissions levels and
comparisons for all vehicles tested, passed, and failed
by station/mechanic, vehicle type, and totals.
82
-3- June 8, 1981
-------
b) Counts, percentages, ratios, emissions levels,
reductions, and comparisons; and types of voluntary
repairs for vehicles which failed the initial
inspection by station/mechanic, mechanic type (licensed
at same inspection station, other licensed, or non-
licensed), vehicle type, and totals.
c) Adjustment and repair costs for vehicles which failed
the initial inspection by mechanic type and totals.
d) Average pass/fail rates and accompanying emissions
levels by vehicle type, reason for failure,
station/mechanic, and totals.
e) Counts, percentages, ratios, and emissions levels,
reductions, comparisons for vehicles which failed the
intital inspection by vehicle type and totals.
f) Inspection, adjustment, and repair costs by vehicle
type, station/mechanic, and totals.
g) Inspection, adjustment, and repair costs and failure
rates by reason for failure, station/mechanic, and
totals.
h) Comparisons of current data with historical data for
trend identification.
2.5.6 Output Descriptions
Although no final decisions have been made regarding the data
output, it appears that there will be a monthly report which will
contain tabular data output, a quarterly report comprised of
aggregated monthly reports and some analysis of the data, and a
formal annual report which will contain extensive analysis and
possible recommendations for program changes. These three products
are more fully described below.
Monthly Report - This report will probably not be distributed
outside of the Department of Health. The purpose of the report will
83
-4- June 8, 1981
-------
be to monitor the I/M program. It will contain little or no
analysis and will probably consist of the following format:
I. The following are tabulated raw data categories:
a) For all vehicles inspected and for each of the
following subcategories:
• Vehicles which pass the first inspection.
• Vehicles which fail the first visual
inspection (by item, e.g., "missing catalyst"
and total).
• Vehicles which fail the first emissions test
(by HC, by CO, both and total)
• Vehicles which fail only the visual.
• Vehicles which fail only the emissions.
• Vehicles which both visual and emissions.
The following tabulations are required:
• Number
• Percent of total failed
• Average emissions
• Inspection cost
• Adjustment cost by:
Vehicle type (technology group, e.g.,
1968-71; model year, make, number of
cylinder)
• Station/mechanic
• Totals
b) For vehicles which failed the initial inspection
(visual, emissions, visual and emissions, visual
only, emissions only, both) and;
• Passed the visual retest
• Passed the emissions retest
• Passed both tests
84
-5- June 8, 1981
-------
• Failed the visual retest
• Failed the emissions retest
• Failed both tests.
The following tabulations are required:
• Number
• Percent
• Average initial and retest emissions
• Emissions reductions between initial and
retest
• Cost of adjustment
• Cost of labor for repairs
• Cost of parts for repairs
(The repair costs should be broken out by
category: tune-up, carburetor, air cleaner,
choke, and other)
by;
Repair/adjustment location (same station
as inspection, other licensed mechanic,
other non-licensed person)
Station/mechanic
- Vehicle type
- Totals
II. The following derived comparisons will be presented in
the monthly report:
a) For vehicles which passed the initial inspection
(both emissions and visual) compare the average
emissions with all inspected vehicles by vehicle
type (technology group, model year, make, number
of cylinders) station/mechanic and totals.
b) For vehicles which failed the initial visual
inspection (by item, e.g., catalyst, and total) or
the emissions inspection (by HC, CO and total)
comparisons with the average emissions of all
vehicles inspected, all vehicles passed by vehicle
type (technology group, model year, make, and
number of cylinders) station/mechanic and total.
85
-6- June 8, 1981
-------
c) For all vehicles which failed the inital
inspection and were adjusted and retested, the
before and after adjustment average emissions by
vehicle type, location, station/mechanic and
totals.
III. For selected raw and derived items comparisons with
equivalent data from previous periods for trend
identification.
IV. Basic statistical analyses of certain raw and derived
data items.
Quarterly Report - This report will probably contain each
of the above mentioned items aggregating over a three month
period. This report may be circulated outside of the Department to
other intested parties (e.g., EPA) and perhaps the general public.
The data in item (d) above (failure rates by model year) will be
used to "fine tune" the cut points such that the state statutory
ceiling on failure rates is not exceeded. Another important item
which must be monitored on a more frequent than annual basis is
item(g). These inspection station data can be used by Department of
Revenue for surveillance and quality assurance. For example, if
certain stations are failing or passing a disproportionate
percentage of vehicles they will be targeted for close
surveillance. The cost data may indicate stations which are
exceeding the statutory cost ceilings for inspections and
adjustments. . , ,
Annual Report - The Air Quality Control Commission is
required under CRS 42-4-309 (9)(a) to submit a report annually to
the Colorado General Assembly evaluating the overall program. This
report, which will be drafted by the Mobile Sources Section, must
include data on the costs and effectiveness of the program. The
report must specifically indicate1 the number of vehicles which do
not met the cut points after an adjustment has been made. This
report will include considerably analysis based upon the monthly and
quarterly reports.
86
-7- June 8, 1981
-------
2.5.7 Quality Control
This section discusses quality control for the
inspection/maintenance data base. A number of checks for invalid
data are possible; and these checks fall in several categories, as
is indicated below.
- •• • s.c i -
A printout should be produced by the computer of any invalid
data identified through automatic checks. Then the errors should be
corrected if possible, using the original report form to compare the
entries in the boxes above and the circles filled in the column
below.
(1) Does the entry have a feasible and reasonable value?
A large number of infeasible or unreasonable
values may be eliminated by coding the data on forms
(Exhibit 2.5-2) with a fixed number of columns and only
certain allowable entries per column and by reading the
data with an optical scanner. For example, it is not
possible to enter the impossible emission value, 150
percent CO, on the form. An automatic or manual check
would be beneficial, however, to determine whether no
circles or more than one circle were filled in under the
same column. Also, the date of test should be earlier than
the date on which the data base quality control check was
done; this check can be done automatically.
(2) Are the different entries on a single form consistent?
Any consistency check which might reveal an invalid
data entry should be made. It should be possible to
automate these checks. The following is a suggested list
of consistency checks:
(a) Is the model year consistent with (i.e., no more than
one greater than) the year of the test?
(b) The certificate status should be consistent with other
entries. For example, if the status is "COMPLIANCE,"
the first or second test emission levels should be
within the standards and the first or second test
visual inspection should have been passed. If a "home
adjust" was performed, requiring compliance with the
standards and not only with the manufacturer's
specifications, then the "home adjust" Pass/Fail
status should be consistent with the reported emission
levels and applicable emissions standards.
87
-8- June 8, 1981
-------
(c) If the catalytic converter or fuel filler neck
restrictor visual inspection was failed, then the year
and model of car should be likely to have a catalytic
converter.
2.5.8 Categorization of Data
Following are the categories (and in one instance,
subcategories) and specific elements of raw data collected. None
are derived data. In parentheses is the rough estimate of the total
number of elements which are expected to be processed each .year
based on 1.5 million vehicle inspections in 1982.
1) Vehicle Identification
• License number (1.5 million)
2) Vehicle Characteristics
• Model year (1.5 million)
• Make (1.5 million)
• No. of cylinders (1.5 million)
3) Administrat ive
• Date of test (1.5 million)
• Inspection station number (1.5 million)
• Inspection station number-for adjustment only (0.1
mi 11 i o n)
• Mechanic's number (1.5 million)
• Mechanic's number-for adjustment only (0.1 million)
• Certificate issued (1.5 million)
4) Cost
• Inspection cost (1.5 million)
• Adjustment cost (0.5 million)
• Repair (labor) cost (0.1 million)
88
-9- June 8, 1981
-------
• Repair (parts) cost (0.1 million)
5) Inspection
Initital Test Visual Inspection (subcategory)
• Presence of catalytic converter (1.5 million)
• Presence of fuel filler neck restrictor (1.5 million)
• Presence of air system (1.5 million)
• Integrity of exhaust system (1.5 million)
Initial Test Emissions (subcategory)
• CO percentage (1.5 million)
• CO pass/fail (1.5 million)
• HC parts per million (1.5 million)
• HC pass/fail (1.5 million)
Retest Visual (subcategory)
• Presence of catalytic converter (0.5 million)
• Presence of fuel filler neck restrictor (0.5 million)
• Presence of air system (0.5 million)
• Integrity of exhaust system (0.5 million).
Retest Emissions (subcategory)
• CO percentage (0.5 million)
• CO pass/fail (0.5 million)
• HC parts per million (0.5 million)
• HC pass/fail (0.5 million)
Repair
Voluntary Repair
• Tune-up (0.1 million)
• Carburetor.(0.1 million)
89 ;
-10- June 8, 1981
-------
• Air cleaner (0.1 million)
• Choke (0.1 million)
• Other (0.1 million)
Home adjust (subcategory) (0.1 million)
• CO percent (0.1 million),
• CO pass/fail (0.1 million)
• HC PPM (0.1 million)
• HC pass/fail (0.1 million)
2.5.9 Recommenda t ions
Currently, the I/M program will be administered jointly by
the CAPCD and the Department of Revenue. As described above, the
CAPCD will gather considerable data from inspection stations on
forms which are read by an optical scanner. The Department of
Revenue, which will be required to gather operating and surveillance
data on a periodic basis, should also report its data (e.g.,
condition of analyzers) by inspection station. Indeed, .the
integrity and reliability of the data gathered by CAPCD from the
inspection forms is dependent, to a large extent, on the accuracy of
the analyzers and the integrity and ability of the mechanics.
Station-specific data reported by the Department of Revenue can be
correlated to the CAPCD data to enhance surveillance and consumer
protection and to identify emissions data which may be inaccurate.
As described in this section, the I/M data to be captured,
stored, sorted and reported relates to the programmatic requirements
of the I/M program. It is possible that the CAPCD Mobile Sourcess
Section will perform FTP testing on a representative sample of
; ,-
vehicles subject to the I/M program to determine the deterioration
associated with adjustments. These data could be stored in the data
base system and used as emission factors, or, to revise I/M credits in
the MOBILE 2 program. The FTP data can also be used along with the
tailpipe emission reductions to determine program effectiveness.
90
-11- June 8, 1981
-------
2.5.10 References
1) Senate Bill No. 52, General Assembly of the State of Colorado,
(Colorado Revised Statutes 42-4-306.5 et seq). Enacted in
1980.
2) Regulation No. 11, "Governing the Motor Vehicle Emissions
Inspection Program for the Control of Air Contaminant and
Emissions from Motor Vehicles," Colorado Air Quality Control
Commission, adopted November 20, 1980.
3) "I/M Data Handling Guidelines Development," work in progress by
Radian Corporation for the U.S. Environmental Protection
Agency. These guidelines will suggest data which should be
generated, captured and analyzed to support local I/M programs.
91
-12- June 8, 1981
-------
3.5 VEHICLE INSPECTION AND MAINTENANCE SUBSYSTEM
3.5.1 Introduction
The Vehicle Inspection and Maintenance Subsystem will consist
of four major processing modules. Each of these modules are made up
of a series of programs or manual procedures normally performed in
one continuous flow to perform a specific task. The modules
included in the Vehicle Inspection and Maintenance Subsystem are:
• Data Collection and Preparation Module
• Master file update module
• On-line error correction module
• Report preparation module
Exhibit 3.5-1 following this page illustrates the data flow for this
subsyst em.
3.5.2 Inputs
3.5.2.1 Vehicle Inspection Forms
The primary data input to the I8cM Sybsystem is from the
vehicle inspection forms. These forms are hand marked by the
inspectors and then forwarded to CAPCD for processing. A sample
inspection form is shown at Appendix 3.5-A.
All other input to the Subsystem is in the form of
corrections to input forms or control and selection data
provided to the report preparation module. These inputs will be
detailed Section 3.5.5.3.
3.5.2.2 Volumes
The volume of data to be processed will be large. If
each of the 1.5 million vehicles in Colorado is inspected, there
will be 1.5 million hard copy forms collected annually.
However, not all data elements will be completed on each form
92
-1- June 8, 1981
-------
VEHICLE INSPECTION AND MAINTENANCE SUBSYSTEM
SYSTEM DATAFLOW
Formt
Colttction
10
oo
Son and Edit
i
r
Updatt Itaur FHa
m
X
5
H
W
-------
because vehicles which "pass" the visual and emissions test are
not required to be adjusted and retested. Only about one-third
of all vehicles are expected to fail, thus two-thirds (or about
1 million forms) will be partially completed. The type of
adjustment or repair and where the repair or adjustment is
conducted will determine whether or not other elements are
entered on the form.
. 3.5.2.3 Data Types
The data collected by this system will be divided into
five categories. The categories of data include various data
elements that will be defined in detail in the data
dictionary. the categories include:
• Administrative and control information
• Vehicle characteristics
• Cost data
• Inspection and emission measurement data .
• Repair and emission measurement data
3.5.3 Outputs
The outputs and reports from the Vehicle Inspection and
Maintenance System would include those reports necessary for file
maintenance and those extracts necessary for periodic reporting.
Those outputs include:
3.5.3.1 Edit reports
This report would contain the rejected inspection records
in a report format similar to the screen format used in the on-
line error correction program. This report, ordered by vehicle
license number, would allow the clerical staff to review the
94
-2- June 8, 1981
-------
rejected Inspection records, identify the errors, and key
corrections into the on-line error correction system.
The report would display each record, grouped by input
batches, highlight the errors, and provide sufficient space for
over-written notes and corrections. The report would then be
used as input to the on-line error correction subsystem.
3.5.3.2 Update control, reports
Various reports would be produced by the update
programs to provide quantity and quality data to assure a
positive control to the updating of the I&M Master File.
Reports produced by these programs should show batch
control data such as batch numbers and record counts by batch as
well as record counts for the master file before and after
updating and number of records posted.
3.5.3.3 Inspection and maintenance reports
Various reports would be produced on a periodic
basis. These reports would include data such as:
• 1st test pass/fail statistics
• 2nd test - adjustments or repair statistics
• Type repair statistics
All reports would be formated according to either
• Year, make of automobile
• Year, make of automobile, inspection station
All reports would be available for:
• Monthly statistics
• Quarterly statistics
• Annual statistics
• Special reports/on demand statistics
95
-3- June 8, 1981
-------
3.5.4 Data Base Description
The major data files for the Vehicle Inspection and
Maintenance Subsystem would be:
3.5.4.1 Input Inspection and Maintenance Forms File
This file would contain the data collected from the
inspection forms after it has been processed through the mark
sense machine.
The original source documents should be retained, in
batches, until that entire batch is successfully posted to the
master file. Similarly, the tape file received from the mark
sense process, should be retained until all batches on that tape
are successfully posted to the master file.
3.5.4.2 Reject Inspection and Maintenance Forms File
This file would contain the records, grouped by batch,
rejected by the editing program. These records would be in the
same format as the input I&M file, available to be -processed in
random order, keyed on license number. The batches would be
held in the reject I&M file until all records in each batch are
free of edit errors, control totals are balanced, and the batch
is released for re-processing by -the master file update
subsystem.
This file is resident on a direct access device and
accessable on-line, in random order, by the error correction
subsystem. Following an error correction cycle, batch records
are processed sequentially, by the edit program.
3.5.4.3 Inspection and Maintenance Master File
The I&M master file would be a complete accumulation of
all the inspection data collected year-to-date. As batches are
released by the edit program, they are sorted, reformated,
updated with the Vehicle Identification Number (VIN) from Motor
96
-4- June 8, 1981
-------
Vehicle files and merged into the I&M Master File in VIN
sequence.
This file would be a sequential file, sorted by VIN, and
stored on tape. The file could be loaded to direct access media
for some short-term purpose, but for most reporting purposes,
the file would be extracted and sorted from the sequential media
in preparation for reporting.
3.5.5 Processing Modules
In this section we will present a discussion of each of the
major functions found in each processing module.
3.5.5.1 Data Collection and Preparation Module
3.5.5.1.1 Forms collection and batching - Manual
activities would be necessary to provide clean,
orderly forms collected and arranged in batches for
input to the mark sensing process. In order to assist
the clerks in this data preparation effort, one corner
of the form should be notched to assist in
straightening the forms. Batch control, consisting of
a batch number and a count of records in the batch,
would also be developed at this time.
3.5.5.1.2 Mark Sense processing - Current plans call for
the batches of inspection forms to be sent to the
Denver Public School District Data Center where they
will be processed and the results returned on standard
data processing tape.
3.5.5.2 Master File Update Module
3.5.5.2.1 Sort and edit- by batch - This program will sort
the input data by batch and record key. The record
97
-5- June 8, 1981
-------
key could be the vebicle license number or other
unique identifier. The editing of each input record
would be by criteria as listed in Appendix E.2
following this section.
3.5.5.2.2 Update master file - The update program would
format the data as required for the Inspection and
Maintenance Master File and merge the records onto
that file. At this point it would also be possible to
add the VIN number from the State Vehicle License
System if necessary.
3.5.5.3 On-line error correction module
3.5.5.3.1 On-line correction program - This program,
operating through CICS, would allow the on-line
correction of errors as identified in the edit
reports. This on-line subsystem would allow access to
the rejected records in a random order and allow the
correction of these records as the information became
available to the clerk. It would also be possible to
arrange these errors into batches to allow for
appropriate controls of throughput by the clerical
staff.
3.5.5.3.2 Select corrected records - This program would
select records that had been corrected and released by
the clerical staff and prepare them for input to the
Master File Update Subsystem.
3.5.5.4 Report Module
This subsystem would consist of from three to five
"sort and report" programs to allow for the periodic reporting
of the statistics gathered by the Vehicle Inspection and
98
-6- June 8, 1981
-------
Maintenance System. Through parameters external to the
programs, the same programs could be used for reports on a
monthly, quarterly, annual, or on-demand basis. These reports
would include:
• Reports ordered by year, make of automobile
• Reports ordered by year, make, station
• Special reports and extracts
99
-7- June 8, 1981
-------
APPENDIX 3.5-A
Edit Criteria
Data input from the Inspection and Maintenance data input
forms should be reviewed by the edit program for reasonableness,
data validity and data consistency.
• Reasonableness checks?
License numbe.r should be checked for conformance
to one of the standard license number patterns
(i.e., XX-NNN).
Station number should be numeric and greater than
zero.
Inspector number should be numeric and greater
than zero.
Model year should be numeric.
Number of cylinders should be numeric.
First test emission levels should be numeric and
greater than zero.
Retest emission levels should be blank or numeric
greater than zero.
Adjustment only station number should be blank or
numeric greater than zero.
Adjustment only inspector number should be blank
or numeric greater than zero.
• All data should be tested to the extent possible for
validity:
Date of test should be of valid month, day, year.
Station number should be from the list of valid
stations.
Model year shuld not be greater than the current
year plus one.
1068- June 8, 1981
-------
Visual inspection failed categories should be
checked against make and model data to determine
if those components were available as original
equipment on that automobile.
Certificate status of "compliance" the emission
levels for either the first test or the second
test should be within the minimum standards for
emissions for that model year automobile.
Consistency checks should be made to ensure that it is
possible to have what otherwise appears to be
reasonable or valid data such as:
The date of the inspection should be prior to the
current date.
Cost amounts for repair should be equal to zero
unless the first inspection test emission levels
exceeded the emissions standards.
101 -9- June 8, 1981
-------
APPENDIX B
103
-------
The following are documents pertaining to statistical analyses and
sampling techniques.
1. Memo from Rick Fawcett to Jerry Gallagher, 8/5/80
2. Memo from Elmer Remmenga to Jerry Gallagher, 8/27/80
3. Report by Elmer Remmenga, 10/9/80
105
-------
c5u
Colorado State University
Department of Statistics 0"'"8' Col°rad°
MEMORANDU
August 27, 1980
TO: Jerry Gallagher, Industrial Sciences
' > I'
FROM: Elmer Remmenga, Statistics '- -'
Survey costs are such that usually a small portion, a sample, of a
population is studied. A cost analysis based on sample size must be
balanced against the precision of the sample results. Samples are
always uncertain because of the vagaries of sampling—randomness,
chance, etc. Intuitively, and theoretically, large samples are more
reliable; but also more expensive. The amount of uncertainty depends
on the inherent variability in the population. This variability can
be determined by theoretical concepts based on mathematical formula-
tions of the distribution (normal, binomial).
As an illustration, consider the probability of heads from tossing a
coin. The coin may be honest or biased. An infinite number of tosses
exist and only after evaluating several million tosses can the probability
of heads be stated precisely. Consider tossing the coin ten times, a
sample of outcomes. If the coin is honest, five heads are expected; but
four or six heads is not unusual and three or seven are not disturbing.
However, at some point (1, 2-, 8, 9 heads) the coin becomes suspect. It is
known that 0 or 10 heads are possible in any given small sample and that
the next small sample will give a different result with larger samples
being more reliable; that is, we feel more secure, more confident, with
the data from a large sample (100 tosses of the coin). Uncertainty is
the name of sampling, thus a sampling strategy is needed.
This paper is a simple look at sampling strategies involving the normal
approximation of the binomial, looking at a range of sample sizes, their
theoretical precision, and their cost.
The normal distribution is the error model most widely used. Here it
provides the multiplier 1.96 which is the upper and lower bound of the
central 95 percent of the area under a normal curve. If a data set is
adequately normal, then 95 percent of its area is bounded by the interval:
the mean ±1.96 standard deviations.
The binomial distribution recognizes only two outcomes (heads or tails,
pass or fail). When the sample size is large (over 30) and the distribution
symmetrical, the discrete binomial can be approximated nicely by the normal
and thus take advantage of the property involving the 95 percent interval.
The standard deviation of the binomial is"Vpq/n where p is the probability
of success (pass or heads) and q = 1-p is the probability of failure (fail
or tails). As n becomes large, the standard deviation becomes small and
thus the estimates more precise,
-------
Jerry Gallagher
Page Two
August 27, 1980
A desirable result would be to be 95 percent sure with a narrow, but
affordable, range of uncertainty.
The N-n/N-1 is called a finite population correction and is needed to
compensate for large sampling proportions (when n/N is large, perhaps
greater than .05) and that will be true here. The sampling involves a
large portion of the population. When n/N is small, .01 or .001 (when N
is large, 1,000,000 or more), then the correction approaches one and is
ignored. Here, it has a real impact with N=100.
EERrlmp
108
-------
SAMPLE SIZE AND SAMI'LH PLAN.CnNSIDrRATIOWS FOR EVALUATING EMISSION TEST
RECORDS FOR CONSUMER PROTECTION
Given 1.4 x 10 test reeor<.l.> each year from 1200 stations at an average
of 100 records per station per month (1200 x 12 x 100 = 1,440,000). Few
test stations will have less than 20 tests per month but some may exceed
500. This range in volume should be considered.
The following suggested sample schemes are based on several assumptions:
-fquality equipment everywhere, equally competent personnel, and random
distribution of automobile.0, partially accomplished by make-model speci-
fications for emission levels. Volume is still a problem.
Based on the 1979 report, a 35 percent decrease in HC can be detected
with 30 to 100 records and a 35 :^jrcent decrease in CO can be detected
with 20 to 25 record.'-:. Thi.~ Lita is somewhat suspect, however, because
it does not really parallel the proposed field testing program and
apparently contains a variety of ti^ained-untrained personnel. It is the
only information readily available on ppm HC and percent CO.
These cliatif'.eG will o-xur only due to the 85 percent of cars that pass
after repair of civ.- 3T; p-TCt.Tit that fail on the first test, about 30 per-
cent of the total. A . ;n.al! :; tat ion handling only 40 tests per month will
have only 12 records that apply. A 25 percent sample will get 36 records
in the sample each year. The average station would have 90 records sampled
per year.
Following an earlier supijysstion, a sample of about 25 percent of the records
(350,000) is proposed. This can be done in two stages.
I. Sample one half of the stations each month. The 50-50 split can
be accomplished with a random number table or by simply assigning
odd-even license numbers to odd-even months. I see no bias in
sequence of,application and/or assigning of station licenses. New
stations can be .i.Kled to the scheme with no problem. A new 50-50
split can be made each 2 months or the same split maintained. Each
station will be sampled 6 times each year.
Randomly, during a month it is sampled, each station will be
evaluated in total to compare the sample with the entire operation
of the station. This can be accomplished with a random number
table, sampling without replacement. Approximately 100 stations
will be evaluated each month.
II. Test records can be drawn from the sample stations each month by
one of the following schemes.
a. Simple random sample. Take one-half (every other record) of
the stations' records in sequence. The order of appearance
for inspection ought not be biased, but there is likely to be
an end-of-the-month rush and possible bias. This will result
in a self weighting sample and will be easy to handle compu-
tationally. Reliable results will be obtained for the population
but large stations will be over sampled and small stations under
109
-------
10/9/80
EER:lmp
-2-
sampled and will have low precision. Sample size would be
about .25 x l.U x 106 = 350,000.
Obtain a fixed number of observations from each station,
say 50 per station, to correspond with a 50 percent sample
of the average station and necessarily take all records for
smaller operations. The sampling would be accomplished in
sequence by taking every n^h record from a file based on
the monthly volume.
Equal precision would be obtained for every station, unless
there is very low volume and then all data would be used and
the precision would be whatever happens for that station. It
would be necessary to use a different blow-up factor for each
station to get total results for the population, but this is
a simple computer problem.
Sample size would be approximately 1200 x 50 x 6 = 360,000.
With equal sample sizes, it would be very easy to obtain the
probability that each station is out of compliance each
sampled month for percent failed and change in HC and CO.
Optical scanner records leading to an overall 80 - 90 percent
sample would be preferable for efficiency, economics, and
statistics.
110
-------
COLORADO DEPARTMENT OF HEALTH
Division or Section of Data Services
INTER-OFFICE COMMUNICATION
TO : Jerry Gallagher DATE : August 5, 1980
FROM: Rick Fawcett SUBJECT: Auto Emission Sampling
This sampling strategy assumes the primary target population to be the auto test
ing station's monthly data, with a secondary target population being all testing
station's yearly data. The calculations used in this report are based on the
following figures:
Primary population •>?& 100
Secondary population, = 1.4x10'
Testing stations ~ 1,166
X keystrokes/card -^ 50
Cost/1000 strokes ^ 1.25
The user is cautioned that changes in the assumed populations may alter the
validity of this sampling strategy.
The primary target population may be sampled with a systematic sampling strategy.
if the collected station data is in random order, the first (n) cards may be se-
lected for each station. If the cards are ordered when received from the testing
stations, the sample must be selected randomly. Random sampling may be achieved
by numbering each card and selecting (n) cards by using a table of random numbers.
The strategy used to select the data cards should be checked for validity with a
statistical comparison of the sample to the population. This check should be
made at least once for each station.
The secondary target population may be sampled with a stratified time sample.
The strata being geographical subsets of the state of Colorado, i.e., counties,
planning regions, or any easily established boundary. This will insure, that on
any given month, the sample will not be geographically biased. The sample time
should be based-.on monthly intervals, so that for any given month a testing sta-
tion will either be sampled or not. Therefore, the time sample selected must be
a factor of 12.
For example, a 50% sample could be drawn from the secondary population by choos-
ing 1/2 of the testing stations from each planning region for 6 months, and the
other 1/2 of the testing stations for the other 6 months.
An alternating month system would reduce the chance of introducing a sampling
bias caused by changes occurring during the ye^r. It would also reduce the
period that a given testing. station is not checked for accuracy.
The sample sizes (n) for the primary and secondary target populations must be
selected to produce an error figure that will be appropriate for the cost, both
in effort and dollars, if a wrong decision is made from the sample statistics.
Formulas and examples are provided which produce an error figure for a given
size. This error factor can be used to determine the range in which the samples
proportion will fall around the population proportion. The error figure is
dependent on the sample si_ze, and may be reduced by increasing the sample size.
.,, Signature
AD BUS-29 (10-29-100)
-------
P.age 2
To: Jerry Gallagher
From: Rick Fawcett
Date: August 5, 1980
Subject: Auto Emission Sampling
Given the differences between the sizes of the primary and secondary target popu-
lations, the error figure for the secondary target population will always be less
than for the primary target population. Therefore, an acceptable error figure at
the primary level will also produce an acceptable error figure at the secondary
level.
Given the cost/1000 keystrokes and the X strokes/card, the data entry costs may be
estimated for different sample sizes.
"2
KS
= population
= primary sample
= secondary sample
= X keystrokes per card
NxKSxn-,xn^ x 1.25 = cost
IxlO3
Population
1.4xl06x50 x 1.25 = 87,500.00
1x10
3
n1=50% n2=50%
(1.4xl06)x50x.5x.5 x 1.25 = 21,875.00
1x10-
n1=30% n2=50%
(1.4xl06)x50x.3x.5
IxlO3
x 1.25 = 13,125.00
The following sampling statistics were prepared by Bruce Ellis, Statistical Analyst
for Health Services.
112
-------
Page 3
To: Jerry Gallagher
From: Rick Fawcett
Date: August 5, 1980
Subject: Auto Emission Sam-
pling
Using a total population size of 100 (N) from a target population defined as the
number of auto emission inspections of any given solution and of which a propor-
tion (P) represents passed inspections, an error term (e) can be calculated for
any given sample size (n) and acceptable confidence level (<*) such that P can be
expected to differ from the sample proportion (p) by no more than 'e' with 100
(1-aO % confidence. 'e1 is calculated as the product of the standard error of p
and the 'z1 value appropriate to the selectedix. (in this case, x/2, applying a
two-tailed ^distribution):
The sample proportion p may be substituted for P, however, since p is also unknown
at this point a value of 0.25 can be used for the quantity £P(1-P)J , since this
is it's maximum value for all values of P between zero and 1.00. Consequently, the
derived e represents an upper bound also. This formula also employs the correction
factor (N-ii) for a finite population, assumed to be roughly 100 for any tested in-
(N-l)
spection station. The e*. level was set at 0.05 (95% confidence), for which a z.
value of 1.96 (two-tailed) was identified.
For n=10, e =
) = .MSb
For n=20, e
/
Q, lo
•ML V J
|i
-------
Page A
To: Jerry Gallagher
From: Rick Fawcett
Date: August 5, 1980
Subject: Auto Emission Sampling
For n=50, e =
/.96
o~o
/oo-5-o \
— J =
/LO •" i -^
These error terms represent an upper bound for all values of P. If P equals 0.65,
e can be calculated for sample sizes 10, 30 and 50 at .2820, .1440 and .940, re-
spectively. Nonetheless, 95% of samples of size 10, drawn from a population of 100
with P=.65, can be expected to derive sample proportions ranging from (P-e) to (P+e),
or fron .3680 to .9320. Consequently, a sample size of less than 30 for a target
population defined as the vehicles inspected by a given station in a given month
seems inadvised.
Should the target population be defined to include all vehicles inspected by a given
station throughout the year, the expected error can be calculated (for an estimated
of 12 x 100 or 1200) for a combined sampling taken every-other month:
For n=60 (10 per month for 6 months), e =
For n=120 (20 per month for 6 months), e =
For n=180 (30 per month for 6 months), e
For n=240 (40 per month for 6 months), e =
J . «1 U
- /
114
-------
Page 5
To: Jerry Gallagher Date: August 5, 1980
From: Rick Fawcett Subject: Auto Emission Sampling
Again, these errors represent upper bounds and may be lower for known values of P.
The critical issue with defining the target population in this manner, however, is
whether the sampling approach is viable, and whether the month and day of the in-
spection represents a systematic bias.
Note: This uses a normal approximation to the binomial distribution. This is
appropriate for sample sizes of around 15 or more if P lies between 0.35 and o.65,
but may be an inadequate approximation for smaller samples.
Source:
Wulpose, R.E. Probability and Statistics for Engineers and Scientists, 1972
pp 201-205
Ferguson, G. A. Statistical Analysis in Psychology and Education. 1971
pp 135-143
115
-------
APPENDIX C
117
-------
The following text is the section of the Vehicle Emissions Control
Study Guide - AIR Progran (January, 1982) which explains the use of
the vehicle inspection forms. The inspection form discussed is the
green fora put in use in January 1982.
119
-------
VEHICLE INSPECTION FORM
General Instructions
The following pages contain samples of the step-by-step procedure
for using and filling out the emissions inspection form. A typical
vehicle is used for the example and hypothetical vehicle data is
presented. For actual inspections the vehicle data should be taken
from the vehicle registration. Actual station and emission mechanic
numbers should always be used. The format used in this presentation
may be different from what is used in real emissions testing. However
if the general procedures used in filling out the form are understood,
no difficulty should be experienced in using a similar form of this
type.
Each form consists of a top sheet and two copies. In addition,
the lower half of copy #3 is the owner's notification of failure to
pass the initial inspection. The top half of copy #3 is the emissions
certificate and should be given to the owner when the inspection or
inspect ion/readjustment process is completed.
Only a number 2 (No. 2) lead pencil may be used in filling out the
vehicle inspection form. Write on a hard surface (clipboard,
tabletop, etc.) and bear down firmly so the recorded information can
be read on all copies. Use care so that the form-is not soiled,
folded, torn, punctured or mutilated. The top (#1 copy) will be
optically scanned by a computer reader. An agent from the Colorado
Department of Revenue will collect the top (#1 copy) periodically.
17e
121
-------
Once the inspection form is Initiated, the official inspection process
has begun and that form cannot be crossed out, erased, or used for any
other vehicle. Inspection forms may be obtained from the Colorado
Department of Revenue and available to only licensed AIR emission testing
stations.
VEHICLE INSPECTION FORMS
I. FIRST TEST INFORMATION
The following instructions show how to complete sections of the form
for vehicles being tested for the first time. Information requested in
blocks #1-9 must be accurately recorded on the form for all vehicles being
inspected. Blocks #10, 11, 16 and 17 must be completed for all vehicles
that pass the first inspection.
A. Obtain the vehicle identification number (VIN) from the registration.
Compare with the VIN on actual vehicle (numbers must agree) and write this
number in block #1, "VEHICLE IDENTIFICATION NUMBER."
Example: ZL45GPR102935.
0VEHICLE IDENTIFICATION NUMBER
122
18e
-------
B. Code the vehicle make In block #2. Example: AMC
C. Write In and code the actual vehicle
(plate and registration must agree) in block
©©©©•©
©©©©©©
©©©©©<
00000©
T) (7) O (T) (T) (7
00000©
0®®©®©
©©©©©©
©©©©©©
©000©©
©0®0©0
©©©©©©
I (5) (*)(?)(*)(?)(*
19e
123
license (metal) plate number
#3. Example: MA1024
If the number has less
than six digits, shift
the number to the right
and code zeros in the
empty spaces to the left.
Temporary, paper license
plate numbers are not to
be recorded.
-------
D. Write in and code the date that the inspection form is started in
block #4. Example: January 5, 1982
DATE OF TEST
e a
E. Write in and code the emissions station license number in block #5.
Example: 0043
STATION
NUMBER
OOM 3
©©©©
€>©•©
If the number has less
than four ditits, shift
the number to the right
and code zeros in the
empty spaces to the left.
F. Write in and code your emissions mechanic license number in block #6.
Example: 0315
If the number has less
than four digits, shift
the number to the right
and code zeros in the
empty spaces to the left.
20e
124
-------
6. Write In and code the last two digits of the vehicle year of
manufacture 1n block #7. Example: 1976.
• C
H. Write 1n and code the number of engine cylinders In block
Example: 1976 AMC, 6 cylinders
i or
en
<0
NOTE: Rotary engine
vehicles should be coded
as to the number of rotors,
I. Write 1n the Inspection cost In block #9. Code the circles under
the dollar portion only.
NOTE: Maximum Inspection
cost 1s ten dollars.
21e
125
-------
Having accurately completed all required information in blocks #1-9, you
are now ready to perform the actual emissions Inspection.
Important; Check the exhaust system for excessive leaks and code the
form as required.
IMPORTANT!
IF EXHAUST SYSTEM
LEAKS ENOUGH TO
CAUSE POSSIBLE
DILUTION OF THE
SAMPLE. FILL IN
THIS CIRCLE. x-v
NOTE: If excessive leaks
are found, proceed with
the Inspection.
II. IDLE TESTS
A. Code the appropriate circle in block #10. Visual inspections are
to be done on ONLY 1982 and newer vehicles. Code "81 or older" if
applicable.
FIRST TEST VISUAL
82 OR NEWER Q
PASSED FAILED
© CATALYTIC CONVERTER ©
© FUEL RESTRICTOR ©
© AIR SYSTEM ©
• 81 OR OLDER
22e
126
-------
B. Following approved procedures, obtain the idle CO and HC readings.
Compare with the maximum allowable emissions standards schedule for CO and
HC; write in and code the actual idle CO and HC readings in block #11,
"FIRST TEST EMISSIONS LEVELS." Example: 2.1% CO and 110 ppm HC.
FIRST TEST EMISSIONS LEVELS
SCO
a1,
Cetteare
tlMM
tethe
MC MMMI
year
tested.
• ©©•
©••O
©0®
©0©
co,
PASS
FAIL
. HC
EMISSIONS STANDARDS
VEHICLE
YEAR
1968-71
1972-74
1975-76
1977-78
1979 6
LATER
CO
7.0%
6.0%
5.5%--
35%
2.0*V
HC
1200ppm
1200ppm
SOOppm
500ppm
4UOppm
C. If CO and HC readings are equal to or less than the applicable
standards, code "COMPLIANCE" in block #16; record the sticker/certificate
number, and sign the inspection form in block #17.
CERTIFICATION ISSUED
COMPLIANCE •
ADJUSTMENT O
DENIED Q
© I certify that I have performed this inspection in accordance
with the roles end guidelines of the Colorado AIR Program.
SIGNATURE OF LICENSED EMISSIONS MECHANIC
The emission inspection is now completed. Forms can be separated and
distributed.
23e
127
-------
III. VEHICLES WHICH FAIL THE FIRST TEST
The following Instructions show how to complete sections of the
Inspection form for vehicles which fall the first test with CO, HC, or both
emissions levels greater than applicable standards. At this time, the
vehicle owner has authorized the original emissions mechanic to make
corrective readjustments to the failed vehicle. (See Sections IV and V for
adjustments made by other than the original emissions mechanic.)
REVIEW
A. Information requested in blocks #1-9 must be accurately completed
and coded as described in Sections I and II.
B. Code the appropriate circle in block #10. Visual inspections are
to be done on ONLY 1982 and newer vehicles (See Section VI). Code "81 or
older" if applicable.
C. Following approved procedures, obtain the Idle CO and HC readings;
compare with the allowable limits schedule for CO and HC readings; write in
and code the actual idle CO and HC readings in block #11, "FIRST TEST
EMISSIONS LEVELS." Example: 1976 vehicle, 7.5% CO and 650 ppm HC.
n) FIRST TEST EMISSIONS LEVELS
•A CO
71*
•©•©
•
ttest
livili
t*th«
stantertt
Mind.
CO
(y PASS
• FAIL
p»«HC
, HC
EMISSIONS STANDARDS
MODEL
YEAR
C0(%)
HC(ppm)
1968-
1971
7.0
1200
1972-
1974
6.0
1200
1875-
1976
5.5
800
1977-
1978
3.5
500
1979*
NEWER
2.0
400
If the CO, HC, or both
readings are above
applicable standards,
notify the owner that the
vehicle has failed the
first test.
24e
128
-------
D. In many cases you will be requested to make the required five
parameter adjustments (Chapter C, Study Guide) and complete the inspection,
After doing the adjustments, retest the vehicle to obtain another set of
final emissions readings. Write in and code the information in block #15,
"FINAL TEST EMISSIONS LEVELS." Example: 0.1% CO and 20 ppm HC.
©FINAL TEST EMISSIONS LEVELS
SCO
OOl
O-O
Crapart
Itvils
lotto
•Missions
sultan's
formM
yaar
taitart.
iHC
002
©000
©©•©
©®©
©0©
©®©
©0©
CO,
PASS
FAIL
HC
E. If the final emissions levels are equal to or below the applicable
standards; code "COMPLIANCE" in block #16. If either CO, HC or both are
above the standards, code "ADJUSTMENT." Write in and code the station
license number, your mechanic license number and the adjustment costs ($15
maximum). Code only the dollar portion in block #12. Sign the form in
block #17.
AIR" PROGRAM ADJUSTMENTS
CERTIFICATION ISSUED
COMPLIANCE •
ADJUSTMENT O
DENIED Q
Sj) I certify that I have performed this inspection in accordance
with the rules and guidelines of the Colorado AIR Program.
SIGNATURE OF LICENSED EMISSIONS MECHANIC
The emissions inspection is now complete. Forms can be separated.
25e
129
-------
IV. VEHICLES THAT HAVE FAILED THE FIRST TEST AND ARE READJUSTED BY
OTHER THAN THE ORIGINAL EMISSIONS MECHANIC
In some cases, the vehicle owner may choose to take the failed
vehicle and attempt to repair or adjust 1t himself, or perhaps to a
non-licensed mechanic. If the owner chooses to do this, separate
the lower half of page #3 (do not separate the form pages) and give
1t to him/her explaining that he/she has up to ten days (from the
date of the first test) to return for one free relnspectlon at your
station.
A. When the owner returns, retest the vehicle to obtain
another set of emissions readings and compare with applicable
standards. If the vehicle passes, write 1n and code these readings
In block #15, "FINAL TEST EMISSIONS LEVELS" and code "COMPLIANCE"
In block #16. Sign the form 1n block #17.
©FINAL TEST EMISSIONS LEVELS
%CO
o|i!5
•©•©
©••
Compare
these
levels
to the
emissions
standards
lor model
year
being
tested.
ppmHC
0& 00
©©•
-8.E 8-
® CERTIFICATION ISSUED
COMPLIANCE •
ADJUSTMENT ©
DENIED O
© 1 certify that 1 have parformad thia inspaction in accordance
with tha rtilai and guidelines of tha Colorado AIR Program.
JeJLrX *sT. WSUWMVU/L--
SIGNATURE OF LICENSED EMISSIONS MECHANIC
B. If the returned vehicle fails this retest, the owner may
now allow you to perform the five parameter adjustments. If the
owner wishes to remove the vehicle to have it repaired/adjusted
elsewhere, he may do so. If the owner authorizes you to perform
the adjustments, make the necessary adjustments.
26e
130
-------
1. If the vehicle now passes, write 1n and code these final
readings In block #15, "FINAL TEST EMISSIONS LEVELS," and code
"COMPLIANCE" In block #16.
@ FINAL TEST EMISSIONS LEVELS
%CO
thorn
livtli
to tin
lor
yoor
taiiog
ttittd
ppmHC
O I 5o
©©©©
©®©
©®®
CO,
PASS
FAIL
, HC
CERTIFICATION ISSUED
COMPLIANCE •
ADJUSTMENT Q
DENIED Q
2. If the vehicle falls after your readjustment, write 1n and
code these final readings 1n block #15 and code "ADJUSTMENT" In
block #16.
©FINAL TEST EMISSIONS LEVELS
%CO
©©•
©•(i
••€
•(
©•(£
©•©
•
Compira
ttiix
livili
to tho
ttandirdi
lor modtl
yur
boioi
tutid.
PASS
FAIL
ppnHC
©©©
©•©
>HC
CERTIFICATION ISSUED
COMPLIANCE ©
ADJUSTMENT •
DENIED Q
27e
131
-------
C. Write in and code the station license number and your
mechanic license number. Write in the adjustment costs but code
only the whole dollar amount in block #12. Sign the form in block
#17.
"AIR" PROGRAM ADJUSTMENTS
© I certify that I have performed this iMieetion in accordance
wife the rule* and guideline* of the Colorado AIR Program.
SIGNATURE OF LICENSED EMISSIONS MECHANIC
The maximum AIR
Program cost for
adjustments is $15.
The emissions inspection is now completed. The form may be separated.
V. ADJUSTMENTS BY ANOTHER LICENSED EMISSION MECHANIC
Some vehicles may be adjusted by another licensed emissions
mechanic outside your shop. Such information will appear on an
"adjustment verification" form (below) attached to the lower half,
page 3 of the original inspection form when the owner returns.
This information (station and emissions mechanic license numbers,
and retest CO and HC emissions readings) must be transferred
(written and coded) to pages 1 and 2 of the inspection form.
MANDATORY ADJUSTMENT VERIFICATION FOR-M
OR 1380 (2/821 ' S«ct1on
-------
Roles may be reversed and you could be asked to readjust a
vehicle that has failed elsewhere. You must be sure to write in
and code your station license and emission mechanic license number
as well as the final CO and HC emissions reading on the adjustment
verification form.
VI. VISUAL INSPECTIONS: 1982 AND NEWER VEHICLES
All 1982 and newer vehicles are required to have a visual
inspection of certain specific emissions control equipment as
originally installed by the manufacturer. From the underhood
emissions control label or decal, determine the type of emission
control equipment installed on the vehicle being inspected,
specifically:
- Air pump or air aspirator system
- Catalytic converter(s)
- Unleaded fuel required
If no decal or sticker is present or this specific information is
not available on the decal/sticker, consult an appropriate shop or
emissions control manual.
Visually inspect the vehicle for the following equipment if
originally equipped:
- Air pump or air aspirator system
- Catalytic converter(s)
- Fuel filter inlet restrictor
All parts of the above system are to be installed and intact.
Only a visual inspection is to be done, no functional or
performance tests are to be performed. If missing or damaged
29e
133
-------
equipment/parts are observed during the first inspection, code
block #10 accordingly and proceed with the inspection. You should
take First Test Emissions Levels for CO and HC and record in block
#11. You may not issue an emissions compliance or adjustment
certificate if the vehicle fails this visual inspection.
ij) FIRST TEST EMISSIONS LEVELS
SCO
•
Coapirt
tb*u
(mil
trtlw
ItrmtM
YMf
Mnf
ttstri.
p»»HC
S)\Z,
©•©©
®@@©
©©©
<
(
©<•)(•
©0©
CO,
PASS
FAIL
.HC
FIRST TEST VISUAL
82 OR NEWER 0
PASSED FAILED
• CATALYTIC CONVERTER ©
© FUEL RESTRICTOR •
• AIR SYSTEM ©
OBI OR OLDER
VII. VEHICLE REPAIRS
Although encouraged, repairs beyond the required five parameter
adjustments are NOT REQUIRED.
Note:
As of July 1, 1982, all 1981 and newer vehicles that fail after
adjustments are performed are required to spend up to a maximum of
$100 of emissions related repairs to lower tailpipe emissions.
Repairs need only be performed until the vehicle is in compliance
with the standards or the $100 maximum has been spent; whichever
comes first. This excludes any required repairs as a result of
failing the Visual Inspection ('82 and newer).
30e
134
-------
Repairs to and/or replacement of items covered in the visual
inspection must be completed before any final adjustments can be
performed. For 1982 and newer vehicles, these items include:
- Air pump or air aspirator system
- Fuel filter inlet restrictor
- Catalytic converters)
After repairs and/or adjustments have been made, a final set of
CO and HC readings should be made. Write in and code these
readings in block #15. Example: 0.3% CO and 25 ppm HC.
©FINAL TEST EMISSIONS LEVELS
%~CO I I ppm HC
•••©
©©•©
(£)•(£
•
CO
Compart
th>M
(avail
to tht
•mission*
stinkards
for modal
yair
being
taitad.
PASS
FAIL
0*5"
©©©©
©©•©
©©©
©©•
©©©
©©©
31e
135
-------
Write In the parts and labor costs (for the repairs which were
performed on the vehicle) in block #13. Code only the dollar
portion.
Example:
Parts = $51.22
Labor = 24.52
Total $75.74
i5 EMISSIONS
REPAIR COST
(81 OR NEWER)
OOO
@©©
©@@
©©©
Note: The "Repair
Cost" (maximum $100)
does not include the
inspection or
adjustment fees.
If the CO and HC readings are equal to or below applicable
standards, code "COMPLIANCE" on block #16.
CERTIFICATION ISSUED
COMPLIANCE •
ADJUSTMENT O
DENIED Q
If the readings are still above standards after appropriate
repairs and/or adjustments, code "ADJUSTMENT" in block #16 and sign
the form in block #17.
© CERTIFICATION ISSUED
COMPLIANCE O
ADJUSTMENT •
DENIED O
© 1 certify that 1 have performed this imp •ction in accordance
with the rale* and guidelines of the Colorado AIR Program.
•JJajLi^. ~~3T. I'M fl (L\*b«UA*>
SIGNATURE OF LICENSED EMISSIONS MECHANIC
32e
136
-------
VIII. INSPECTION FORM DISTRIBUTION
After the vehicle Inspection form has been completed or more
than ten days have elapsed since the first inspection, the
inspection form document is to be separated and distributed as
follows:
PAGE #1 - Hold for pick-up by the Department of Revenue
investigator. DO NOT fold, staple, mutilate, soil or
otherwise damage this section.
. PAGE #2 - Retained by the licensed emission testing station for
purposes of record keeping.
PAGE #3 - Vehicle owner's copy.
IX. EMISSION CERTIFICATE: "DENIED"
If a failed vehicle is not returned for a retest within ten
days of the initial inspection, code block #16 "DENIED."
If the vehicle owner refuses to allow a licensed AIR Program
emissions mechanic to perform corrective readjustments and/or the
'81 and newer repairs to a vehicle that fails the retest, code
block #16 "DENIED."
©FINAL TEST EMISSIONS LEVELS
CO
©.(«
••*
(!).(»
CO ,
Campari
these
levels
to the
emissions
standard!
for modal
yaar
being
tested.
PASS
FAIL
ppmHC
0130
. HC
O CERTIFICATION ISSUED
COMPLIANCE O
ADJUSTMENT Q
DENIED •
33e
137
-------
X. FREE REINSPECTION
Each emissions testing station must provide ONE FREE REINSPECTION
upon request for each vehicle originally tested and failed either
for visual or tailpipe test. This obligation extends for a period
of ten days following the initial inspection.
34e
138
-------
APPENDIX 0
139
-------
The following documents are slightly reduced copies of three types
of inspection forms. The first is the "blue form" in use for inspec-
tions from July 1, 1981 through December 31, 1982. The second is the
"green form" in use for inspections conducted after December 31, 1981.
The third is the newest type o^ form and will be used beginning
July 1, 1982. For a discussion of differences between the first two
types of forms, consult section 3.1. The actual forms are 81/fc" x 12"
with a 5/8" strip at the top to hold the three pages and two carbon
sheets together. Page one is white with colored ink. Page two is
light blue with colored ink, page three is yellow with colored ink.
Pages two and three are of lighter weight paper than the
"Trans-Optic"® bond of page one. The instructions (page four) are
printed on the reverse of page three. Alternate columns are shaded
on all forms.
141
-------
BLUE FORM
July - December, 1981
143
-------
NCS Ttana-Optic B10-31366 321
COLORADO DEPARTMENT OF HEALTH
"AIR" PROGRAM INSPECTION/READJUSTMENT
REPORT
©
VEHICLE IDENTIFICATION NUMBER
AUTO
MAKE
AMC ©
AUDI ©
AUHE ©
AUS1 ©
BMW ©
BUCK ©
CADI ©
CHEK ©
CHEV ©
CHRY ©
OATS ©
DODG ©
FIAT ©
FORD ©
HOND©
INTE ©
JAGU ©
JEEP ©
LANC ©
LINC ©
MAZD ©
MEBZ ©
MERC ©
MG ©
OLDS ©
OPEL ©
PLYM ©
PONT ©
PORS ©
PUGT ©
RENA ©
SAAB ©
SUBA ©
TOY ©
TRIP ©
VOLK ©
VOLV ©
QTHR Q
© LICENSE PLATE
©©©©©©
©©©00©
©©©©©©
©©©©©©
©00©©©
©©©©0©
2)©®©®®
©0©©©0
®®®®@®
®0®®0®
®®®®@0
©©©©©©
©0©©©©
O©O©O©
©Q©©©©
©©©©©0
©©©©©©
©©©©©©
©©©©©©
©©©©©©
I/CL/CL/CI
©©©©©©
©©©©©©
©©©©©©
©©©©©©
DATE OF TEST
©
©
©
(T)
3©
©
©
©
STATION
NUMBER
©©©©
©©©©
®@©®
©©©©
©0©®
®@®0
©0Q©
©0©®
. MECHANIC'S
' NUMBER
©0©®
©©©©
©®0©
©©©©
©©©©
®©@®
©00©
®@®0
@®@®
©«a
©0
©
©
CERTIFICATION ISSUED
COMPLIANCE ©
ADJUSTMENT ©
DENIED ©
COSTS (IN DOLLARS)
INSP.
©©
©
©
©
©
ADJ.
©
©
©
©
©
©
REQUIRED REPAIR
LABOR
©0©
©©©
©0©
©0©
©0©
©©©
©0©
©0©
PARTS
©©©
©Q©
©@©
©©©
©©©
©0©
©0©
RETEST
VISUAL INSPECTION
FAILED
CATALYTIC CONVERTER ©
FUEL RESTRICTOR ©
AIR SYSTEM O
EXHAUST
SYSTEM INTEGRITY Q
RETEST EMISSIONS LEVELS
% CO
)©.©
©0.0
Compare
these
levels
to the
emissions
standards
for modal
year
Doing
teslod.
ppm HC
®©0©
©00©
©00
© PASS ©
© FAIL ©
I certify that I have performed this inspection
and any required adjustments in accordance with
the rules and guidelines of the Colorado AIR
Program.
SIGNATURE OF LICENSED EMISSION MECHANIC
124693
EMISSIONS STANDARDS
VEHICLE
YEAR
1968-71
1972-74
1975-76
1077 78
19791.
LATER
CO
7.0%
6.0%
b5%
35%
2.0%
HC
1200ppm
1200ppm
800ppm
SOOppm
400ppm
© FIRST TEST
VISUAL INSPECTION
FAILED
CATALYTIC CONVERTER ©
FUEL RESTRICTOR ©
AIR SYSTEM ©
EXHAUST
SYSTEM INTEGRITY ©
@ FIRST TEST EMISSIONS LEVELS
% CO
0©
Compare
these
levels
to the
emissions
standards
for model
year
being
tested.
ppm HC
©©©©
Q©©0
©©©
®@©
®@®
®@®
®@®
© VOLUNTARY
REPAIR
TUNEUP ©
CARBURETOR ©
AIR CLEANER ©
CHOKE ©
OTHER ©
(specify)
Q HOME ADJUST
% CO
T
PASS
FAIL
ppmHC
0
0
© PASS ©
© FAIL ©
USE A NUMBER 2 PENCIL ONLY ON THIS FORM
FOLLOW INSTRUCTIONS ON REVERSE SIDE
THIS COPY TO DEPARTMENT OF HEALTH
144
ADJUSTMENT ONLY
STATION
NUMBER
©©0©
©00©
©00©
©®@©
MECHANIC
NUMBER
,0)® ® ©
Q©Q©
©00©
©0©0
0©©©
PAGE1
-------
COLORADO DEPARTMENT OF HEALTH
"AIR" PROGRAM INSPECTION/READJUSTMENT
REPORT
o
VEHICLE IDENTIFICATIOM NUMBER
AUTO
MAKE
AMC O
AUDI O
AUHE O
AUST O
BMW O
BUCK O
CADI O
CHEK O
CHEV O
CHRY O
OATS O
DODG O
FIAT O
FORD O
HONDO
INTE O
JAGU O
JEEP O
LANC
LINC
MAZD
MEBZ
MERC
MG
OLDS
OPEL O
PLYM
PONT
PORS O
PUGT
RENA
SAAB O
SUBA O
TOY
TRIP
VOU
VOLV
OTHR Q
© LICENSE PLATE
©0©©©©
©©©©©©
©©©©©©
3)©©©@©
©00©©©
©©©©©©
©©©©©©
©0©©0©
0®®®®®
©©©©©©
©©©©©©
©©©©©©
©
0©O©O©
©©©©©©
©(
©©©©©©
©®©@®®
Z) O O O QL
©0©©©©
©©©©©©
DATE OF TEST
©
©
INSP.
O
EMISSIONS STANDARDS
VEHICLE
YEAR
1968-71
1972-74
1975-76
1977-78
1979 b
LATER
CO
7.0%
6.0%
5.5%
3.5%
2.0%
HC
1200ppm
12DOppm
BOOppm
SOOppm
400ppm
©
5)©
3)©
©
©
©
©
©
©
©
©
STATION
NUMBER
©©©©
©©©©
©©©©
©©©©
©©©©
©©©©
©©©©
. MECHANIC'S
1 NUMBED
©©©©
©©©©
©©©©
®©®0
©00®
©©©©
©©©©
0©
0©
5)©
CERTIFICATION ISSUED
COMPLIANCE O
ADJUSTMENT O
DENIED O
COSTS (IN DOLLARS)
AOJ
©
©
©
©
REQUIRED REPAIR
LABOR
©©©
0©©
©®©
©0©
©0©
©®©
PARTS
©©©
®©0
RETEST
VISUAL INSPECTION
FAILED
CATALYTIC CONVERTER ©
FUEL RESTRICTOR ©
AIR SYSTEM Q
EXHAUST
SYSTEM INTEGRITY Q
RETEST EMISSIONS LEVELS
% CO
©O-©
©O-0
O.®
Coalite
these
to the
emissions
stondords
lor model
f»i
being
teit>d.
ppmHC
©0©©
©©©©
©®0©
©0©
©0®
© PASS
© FAIL
is:
© I certify that I have performed this inspection
and any required adjustments in accordance with
the rules and guidelines of the Colorado AIR
Program.
SIGNATURE OF LICENSED EMISSION MECHANIC
124693
FIRST TEST
VISUAL INSPECTION
FAILED
CATALYTIC CONVERTER O
FUEL RESTRICTOR ©
AIR SYSTEM ©
EXHAUST
SYSTEM INTEGRITY O
© FIRST TEST EMISSIONS LEVELS
CO
Compile
these
llvtls
to the
emissions
itinditdt
for model
yen
being
tested.
©
©
ppm HC
©©©©
©©Q©
©©©©
®@©
VOLUNTARY REPAIR
TUNE-UP
CARBURETOR
AIR CLEANER
CHOKE
OTHER
(specify)
O
o
o
o
o
HOME ADJUST
SCO
»*>HC
PASS ©
FAIL ©
ADJUSTMENT ONLY
STATION
NUMBER
0©0©
©00®
0©©©
MECHANIC
NUMBER
©00©
3)©G
©©©©
©00©
©0©®
©©©©
® ® ® ®
000©
PASS
FAIL
THIS COPY TO BE, FILED AT AIR PROGRAM STATION
PAGE 2
-------
NCS Tran»-Optic BIO 31356 321
COLORADO DEPARTMENT OF HEALTH
"AIR" PROGRAM INSPECTION/READJUSTMENT
REPORT
VEHICLE IDENTIFICATION NUMBER
124693
AUTO
MAKE
AMC O
AUDI O
AUHE O
AUST O
BMW O
BUCK O
CADI O
CHEK ©
CHfV ©
I;HRY O
UATS O
OODG ©
FIAT O
FORD O
MONO O
INTE ©
JAfJU O
JtEP O
LANC O
LINC O
MAZD©
MEBZ
MERC
MG ©
OLDS O
OPEL O
PLYM O
PONT O
PORS O
PUGT O
RENA O
SAAB O
SUBA O
TOY ©
TRIP
VOLK
VOLV
OTHR Q
LICENSE PLATE
LOU
© DATE OF TEST
MO
OAV
VH
1
/TI STATION
^ NUMBER
IMPORTANT INFORMATION TO VEHICLE OWNERS
1. You must keep both halves of this form to present to a safety
inspector at the time of vehicle safety inspection.
2. You must have both halves of this form to transfer to the new
owner should you sell this vehicle.
3. YOU MUST HAVE A CERTIFICATION OF EMISSIONS CONTROL
BEFORE YOU SELL THIS VEHICLE.
4. KEEP THIS CERTIFICATION WITH YOUR REGISTRATION.
0
CERTIFICATION ISSUED
COMPLIANCE ©
ADJUSTMENT O
DENIED O
PLEASE
READ
WARNING
BELOW:
@ COSTS (IN DOLLARS)
INSP
ADJ.
REQUIRED REPAIR
LABOR
PARTS
WARNING:
II the retest visual inspection portion of this form (box 11)
indicates that this vehicle is missing any emissions control
equipment, you may be required to have this equipment
installed before you receive your next emissions inspection
and/or safety inspection. Please contact a licensed emis-
sions mechanic or the Department of Health for specific
information.
RETEST
VISUAL INSPECTION
FAILED
CATALYTIC CONVERTER O
FUEL RESTRICTOR O
AIR SYSTEM O
EXHAUST
SYSTEM INTEGRITY Q
RETtST EMISSIONS LEVELS
v. CO
Compere
these
levels
to the
emissions
standard!
for model
veer
being
tested.
ppmHC
©00
©
©
PASS
FAIL
©
Ul
W!
3
82
85]
© I certify that I have performed this inspection
and any required adjustments in accordance with
the rules and guidelines of the Colorado AIR
Program.
SIGNATURE OF LICENSED EMISSION MECHANIC
124693
MOTE TO VEHICLE OWNER
I You may make an.1 ad
lustmenls or (epairs your
sell or have anyone else
do Idem, bui your vehicle
must meet the emissions
standards or be adjusted
by a licensed emissions
mechanic before a Certi
hcation of Emissions Con
Irol can be issued
2 II the exhaust integrity
portion of the visual m
speutiun (box 9) is check
ed. the exhaust system
must be repaired before
a valid emissions mea-
surement can be obtained
EMISSIONS STANDARDS
VEHICLE
YEAR
19b» 71
1972 74
1975 76
1977 78
1979 b
LATER
CO
7 O1*,
6U%
b5%
35%
2u"X,
HC
UOOppm
UOOppm
eOQppm
500ppm
400ppm
© FIRST TEST
VISUAL INSPECTION
FAILED
CATALYTIC CONVERTER ©
FUEL RESTRICTOR O
AIR SYSTEM ©
EXHAUST
SYSTtM INTEGRITY ©
@ FIRST TEST EMISSIONS LEVELS
••; CO
I
000000000©
©0
Compare
these
levels
to the
emissions
standards
tor moriel
year
being
tested
ppm HC
0000000000
0000000000
© VOLUNTARY
REPAIR
TUNEUP ©
CARBURETOR ©
AIR CLEANER O
CHOKE ©
OTHER ©
(specify)
Q, HOME ADJUST
(see Note to
Vehicle Owner)
% CO
"T
PASS
FAIL
ppmHC
O
O
©
©
© ADJUSTMENT ONLY
STATION
NUMBER
MECHANIC
NUMBER
PASS
FAIL
©
©
IMPORTANT - KEEP THIS FORM
It will allow you to have one Iree re
inspection at the original inspection
station within ten days if your vehicle
fails its first inspection.
FOR ANY ADDITIONAL INFORMATION
CONTACT
COLORADO DEPARTMENT OF HEALTH
Denver: 3204180
Aurora; 364-4135
Fl Collins 221-5324
Colo. Springs.
146
THIS COPY RETAINED BY VEHICLE OWNER
PAGE 3
-------
INSTRUCTIONS TO LICENSED EMISSIONS MECHANICS
(ONLY a Licensed Emissions Mechanic may perform this inspection.)
NOTICE
The first copy (top) of this form will be scanned electronically.
Please keep the top copy free of dirt or grease.
Oo NOT fold, staple, spindle, or mutilate top copy.
Do NOT make ANY marks on the top copy other than specified in directions.
You may make notes on the second or third copies ONLY.
Use ONLY a NUMBER TWO pencil on this form.
Where boxes are provided put one and only one number or letter in each box, if there are more boxes
than numbers or letters put enough zeroes in front (to the left) to fill up the extra boxes. For example,
if the emission levels are 4.8% CO and 750 ppm HC:
ENTER:
00 NOT ENTER:
Except for Item O (Vehicle Identification Number), after the boxes are marked, fill in the circle under
each box which has a letter or number that matches the letter or number in the box above. There
should be just one circle filled in under each box.
FIRST INSPECTION
A. Complete items O through ©
6. Oo the first test visual and emissions inspection © and ©. (mark PASS or FAIL for CO and HC).
C. If the vehicle passes both the visual and emissions inspection, mark "Compliance" ©. sign the form ©, punch | COC |
®. and "Date of Test" on pages 2 and 3. and give both halves of page 3 to the vehicle owner.
D; If the vehicle fails the visual or emissions inspection and the adjustments or repairs are done at your station, complete
item © and go on to Step G.
If the vehicle owner wants the work done elsewhere, give him/her the bottom part of page 3 to be filled in by the person
who does the work and returned within 10 days for a free reinspection.
REINSPECTION
E. If the vehicle was adjusted or repaired by a Licensed Emissions Mechanic, copy the station and mechanic's license num-
bers from the "Adjustment Verification" form under item ©. attach one copy of the "Adjustment Verification" form to page
2. and go on to Step G.
F. If the vehicle was adjusted or repaired by a non-licensed person, complete item ©. If the vehicle passes, go on to Step
G. If it fails, make the adjustments or repairs required and go on to Step G or mark "Denied" ©. sign the form ®.
punch CD @. and "Date of Test" on pages 2 and 3. and give both halves of page 3 to the vehicle owner.
G. If the vehicle failed the first test visual, do the retest visual inspection 0. if it passed the first test visual, go on to step H.
H. Do the retest emissions inspection ©. (mark PASS or FAIL for CO and HC).
I. Record the amount charged for the inspection, the adjustments (if performed), and the labor and parts costs for required
repairs ©
J. Record any voluntary repairs ©.
K. If the vehicle passes both the visual and the emissions retests. mark "Compliance" ©. sign the form ©, punch [COC|
@. and "Date of Test" on pages 2 and 3, and give both halves of page 3 to the vehicle owner.
L. If the vehicle passes the visual retest but fails the emissions retest (and, for 1981's and newer, $100.00 was spent on
emissions repairs) mark "Adjustment" ©, sign the form ®. punch | COA | ©. and "Date of Test" on pages 2 and
3, and give both halves of page 3 to the vehicle owner.
M. If the vehicle fails the visual retest (or. for 1981's and newer, fails the emissions retest and less than $100 was spent
on emissions repairs) mark "Denied" ©. sign the form ®, punch | CO | ®. and "Date of Test" on pages 2 and
3. and give both halves of page 3 to the vehicle owner.
147
-------
GREEN FORM
January, 1982 - Current
149
-------
COLORADO "AIR" PROGRAM REPORT
NCS Trans-Optic B10-32041-321
FINAL TEST VISUAL
82 OR NEWER Q
FAILED
© CATALYTIC CONVERTER
© FUEL RESTRICTOR ©
F> AIR SYSTEM
CHSV O
0
SATS
©FINAL TEST EMISSIONS LEVELS
am O
O
O
IAMC O
im O
0)(0
©00©
Compare
these
levels
to the
emissions
standards
for model
year
being
tested.
LEAKS EJKRWH TO
564450
PROGRAM ADJUSTMENTS
EMISSIONS
REPAIR COST
(81 OR NEWER)
TOY O
TRIP O
VftLK O
O
CATAIYW fiO«¥E8T«ft *
FIRST TEST EWISSKBtS LEVELS
V, CO
I»**U
to the
«t»i«Urd<
b*in»
' W>««e
CO,
PASS
i FAI1
©©©©
©@©©
©00©
©00©
©©©©
©©©©
©0® ©
©0®®
0©Q©
©00©
©00©
0©®®
©©©I
©0©l
©0©l
©001
©©©I
©001
©©
©©©I
©0®l
USE
ONLY A NUMBER TWO (2) PENCIL ON THIS FORM
FOLLOW INSTRUCTIONS ON REVERSE SIDE
150
EMISSIONS STANDARDS
MODEL 1968-
YEAR 1971
C0(%) 7.0
HC(ppm) 1200
1972-
1974
6.0
1200
1975-
1976
5.5
800
1977-
1978
3.5
500
1979&
NEWER
2.0
400
with the rate* «nd »ultfelme* ol tts Ctttrtato A(fl fngtam.
THIS COPY TO BE PICKED UP BY DEPARTMENT OF REVENUE
PAGE 1
-------
COLORADO "AIR" PROGRAM REPORT
NCS Trans-Optic B10-32041-321
:j> •$%?$%& %$j$&£^%K%^^^i^fpl'4v^ v\ ;^v 'x '^v;
BJJPl^g^ - ^
•j--; '•
CAtH
AMlfd
ft«ot O
MM®
AW®
O
Q
mr O
rasa O
HWfflQ
WTI O
JA6UO
46EP O
LANC O
O
PLYM
$AABO
SU8A0
IftY Q
WW O
m
SATtOfTBT
MZ
S> ] REMARKS:
rtWKTUW
«CATAtYT?ECO«tfEftTe g
Qli m ataiB
%co
T
®^@
hoit*
tetW
ttrtt*.
.>>*>«,
151
••»•««:
^vT^ f
,:'Kl-l
.^
..,m.
Z
0_
FINAL TEST VISUAL
82 OR NEWER ©
PASSED FAILED
© CATALYTIC CONVERTER ©
© FUEL RESTRICTOR ©
© AIR SYSTEM ©
© 81 OR OLDER
©FINAL TEST EMISSIONS LEVELS
% CO
Compare
these
level*
to the
emissions
standards
lor model
year
being
tested.
ppm HC
©©©©
©@©0
©0©
©©©
©0©
CO,
PASS
FAIL
, HC
564450
"AIR" PROGRAM ADJUSTMENTS
STATION
NUMBER
©©©©
®®@®
®@®®
©©©©
©©©©
®®0©
®0®0
MECHANIC'S
NUMBER
®@®@
©©©©
©0©©
®®®@
000©
<2 EMISSIONS
REPAIR COST
(81 OR NEWER)
s s s |* |*
©0©l
©©©I
®©®l
®®®l
©0®l
®®®l
®©©l
®®®l
0001
EMISSIONS STANDARDS
MODEL 1968-
YEAR 1971
CO(Vi) 7.0
HC(ppm) 1200
1972-
1974
6.0
1200
1975-
1976
S.S
BOO
1977-
1978
3.5
500
1979&
NEWER
2.0
400
THIS COPY TO BE FILED AT "AIR" PROGRAM STATION
PAGE 2
-------
I I
IM
COLORADO "AIR" PROGRAM REPORT
NCS Trans-Optic B10-32041-321
BMWO
TORS O
RJ8TO
Rlf&Q
SAAB0
StfBAQ
TOY
TRIP
VtJUf Q
V01V O
82 OR NEWER Q
PASSED FAILED
© CATALYTIC CONVERTER ©
RESTRICTOR ©
SYSTEM ©
O 81 OR OLDER
©FINAL TEST EMISSIONS LEVELS
v. co
Compare
thgu
to the
(missions
standards
for model
yeer
being
tested.
ppmHC
(£)(i
©0©
©0©
co,
PASS
FAIL
. HC
FIRST TKT VISUAL
82 m tmm o
I
JCATAlYTlCCftNVarFEft
564450
© "AIR" PROGRAM ADJUSTMENTS
STATION
NOMBER
MECHANIC'S
NUMBER
ADJUSTMENT
COST
$
s
*
t
<3> EMISSIONS
REPAIR COST
(81 OR NEWER)
«*!»
t««l$
twmwltt
»«*•
PASS
NOTE TO VEHICLE OWNER
If your vehicle did not pass the first lest, you may make any adjustments or repairs yourself
or have anyone else do them but. in order to receive a Certification of Emissions Control.
ynur vehicle must meet at least ene of the following conditions:
1) Pass both the visual and the emissions final tests, or
2) Pass the visual final test and be adjusted by a licensed emissions mechanic
and. for 1981 and newer vehicles, have at least one hundred dollars ($100)
in emissions-related repairs done.
fOR ANY ADDITIONAL INFORMATION. CONTACT ONE OF THE FOLLOWING EMISSION
TECHNICAL CENTERS:
Denver Metro Area - 9640 E. Colin Ave.. Aurora 364-4135
Fort Collins Area - 429 N. College Ave.. Ft. Collins 221-S324
Colo. Springs Arei - 1403 S. Tejon St.. Colo. Springs 633-2333
OR THE COLORADO DEPARTMENT OF REVENUE 866-5518
IMPORTANT - KEEP THIS FORM
It will allow you to have one free re-
inspection at the original inspection
station within ten days if your vehicle
fails its first inspection.
EMISSIONS STANDARDS
MODEL
YEAR
C0(%)
1968-
1971
7.0
1972-
1974
6.0
HC(ppm) 1200 1200
1975-
1976
5.5
BOO
1977-
1978
3.5
500
19794
NEWER
2.0
400
arrtt «t» wtes «n4 pitWiw*
-------
INSTRUCTIONS TO LICENSED EMISSIONS MECHANICS
(ONLY a Licensed Emissions Mechanic may perform this inspection.)
NOTICE
• The first copy (top) of this form will be scanned electronically.
• Please keep the top copy free of din or grease.
• Do NOT fold, staple, spindle, or mutilate top copy.
• Do NOT make ANY marks on the top copy other than specified in directions.
• You may make notes on the second or third copies ONLY.
• Use ONLY a NUMBER TWO pencil on this form.
• Where boxes are provided put one and only one number or letter in each
box. if there are more boxes than numbers or letters put enough zeroes in
front (to the left) to fill up the extra boxes. For example, if the emission
levels are 4.8% CO and 750 ppm HC:
• Except for Item 0 (Vehicle Identification Number), after the boxes are
marked, fill in the circle under each box which has a letter or number that
matches the letter or number in the box above. There should be just one circle
filled in under each box.
ENTER
% CO
c
4[ 8
ppm HC
0
7
5
0
DO
NOT
ENTER
FIRST INSPECTION
PRINT - Name and address of vehicle owner at top of this form
ITEM
O Vehicle Identification Number - Write in VIN number.
© Auto Make - Fill in circle for auto make.
© License Plate - Enter license plate number and fill in circles (if
temporary tag leave blank, if more than six characters use first six).
© Date of Test - Enter month, day and year and fill in circles.
© Station Number • Enter your station's license number and fill in
circles.
© Mechanic's Number - Enter your emissions mechanic's license
• number and fill in circles.
© Model Year - Enter last two digits of vehicle's model year and fill in
circles (example: 1972 enter 72).
© No. of Cyl. - Enter number of cylinders in engine and fill in circles
(if rotary engine enter number of rotors).
© Inspection Cost - Enter cost of inspection in dollars and cents
(maximum $10.00) and fill in circles under dollar portion only.
© First Test Visual - For 1981 and older vehicles, fill in "81 OR
OLDER" circle (no visual inspection needed).
For 1982 and newer vehicles, fill in "82 OR
NEWER" circle, complete first visual inspection
and fill in a pass or fail circle for each emissions
control system.
© First Test Emissions Levels - Complete first emissions inspection,
enter readings and fill in circles. Compare readings to State standards
and fill in pass or fail circles for CO and for HC.
NOTE: THIS COMPLETES FIRST INSPECTION. IF VEHICLE HAS
PASSED ALL REQUIREMENTS. FILL IN "COMPLIANCE" CIRCLE IN
ITEM © AND SIGN ITEM © . ENTER CERTIFICATE OF EMISSIONS
CONTROL NUMBER AT TOP OF FORM, GIVE THIRD COPY TO
VEHICLE OWNER AND AFFIX CERTIFICATE OF EMISSIONS
CONTROL STICKER TO THE WINDSHIELD AS SPECIFIED. 153
VEHICLES FAILING FIRST INSPECTION
ITEM
"AIR" Proaram Adjustments - If adjustments were made by a
licensed emissions mechanic, enter license numbers of station and
mechanic where adjustments were made and cost of adjustments in
dollars and cents (maximum $15.00). Fill in circles under station
number, mechanic's number and dollar portion of adjustment cost.
NOTE: If adjustments were made by another licensed emissions
mechanic, copy this information from "Adjustment Verification Form".
' Emissions Repair Cost - For 1981 and newer vehicles, enter cost of
emissions related repairs in dollars and cents and fill in circles
under dollar portion only.
Final Test Visual - For 1981 and older vehicles, fill in "81 OR
OLDER" circle (no visual inspection needed).
For 1982 and newer vehicles, fill in "82 OR
NEWER" circle, complete final visual inspection
and fill in a pass or fail circle for each emissons
control system.
Final Test Emissions Levels - Complete final emissions inspection,
enter readings and fill in circles. NOTE: If adjustments were made
by another licensed emissions mechanic, copy readings from
"Adjustment Verification Form". Compare readings to State standards
and fill in pass or fail circles for CO and for HC.
Certification Issued - Fill in circle next to.-
COMPLIANCE if item © passed and item © is within State
standards.
ADJUSTMENT if item © passed, item © failed, adjustments
done by licensed mechanic and, for 1981 and
newer vehicles, at least $100 spent on
emissions related repairs.
DENIED if item © failed or item © failed and 1)
adjustments not done by licensed mechanic or
2) for 1981 and newer vehicles, less than
$100 spent on emissions repairs.
© Signature of Licensed Emissions Mechanic • After completing this
form, sign it, enter number of the certificate issued (or "none" if
certification denied) at top of form and give third copy to vehicle
owner. If a certification is issued, affix certificate of emissions
control sticker to the windshield as specifier).
-------
NEWEST INSPECTION FORM
For use beginning July 1, 1982
155
-------
COLORADO "AIR" PROGRAM REPORT
NCS Tram-Optic BIO 32551 321
— 0 vaoeu iOEsmtewtoH mm*
so.
Q
O
Q
.O
G
Q
FIAf Q
O
O
4SEP O
um O
MERCQ
MS O
H.VMO
SUBA O
jot O
O
am O
n
€>€>©©®©
®{2>
©0®®®®
JffiL
©
MBMtt
HL
HBPORTAKTt
IF EXHAUST SW«M
THIS CIRCLE.
FINAL TEST VISUAL
82 OR NEWER Q
PASSED FAILED
© CATALYTIC CONVERTER ©
© FUEL RESTRICTOR ©
© AIR SYSTEM ©
OR OLDER
©
FINAL TEST EMISSIONS LEVELS
2500 RPM-
% CO
©©•©
©©•©
©•©
©•©
©•©
©•©
©•©
•81 OR
NEWER
Compiri
these
Isvili
to the
emissions
stendards
lot model
year
being
tested.
ppmHC
©oo©
©©©©
@®©
© PASS ©
C°© FAIL ©HC
SCO
©©•©
©©•©
©«»*
PASS
FAIL
©0-®
(Ml
tlws»
«»«*#*
»•**•
•s
PASS
FAIL
. HC
©
"AIR" PROGRAM ADJUSTMENTS
STATION
NUMBER
©©©©
©OQQ
©©©©
©©©©
©00©
©©©©
©©©©
MECHANIC'S
NUMBER
©©©©
O0O©
©©©©
ADJUSTMENT
COST
© EMISSIONS
REPAIR COST
(81 OR NEWER)
©©©I
©©©I
©©©I
©©©I
©©©I
®®®|
©©©I
®®®|
USE ONLY A NUMBER TWO (2) PENCIL ON THIS FORM
FOLLOW INSTRUCTIONS ON REVERSE SIDE
EMISSIONS STANDARDS FOR THIS VEHICLE
HC
CO
I ctn*t that! tun* fvtttmtt tfci» imtntiw a *ww4»nt»
wirt tte futu Ml gai4*!imi ef tt* Celsufc Alft f>«gr«m.
SI6NATItB< Of LIEtSSEO tSIISSIOSS BECHA»IC
THIS COPY TO BE PICKED UP BY DEPARTMENT OF REVENUE
156
PAGE1
-------
COLORADO "AIR" PROGRAM REPORT
NCS Trans-Optic B10-32661-321
.mftMMfftttftli;
O
Q
wtsrO
BAM Q
GHiK O
CHEV Q
O
O
«AT
fOBB O
HOffflO
am O
> 0
uwe O
WK'Q
we -0
Oi»$ O
LfL
SCO
X£>0
&@<£)®
JD
©
ttttMACtt
REMARKS:
fjRSTTiSt WSOAL
Q»t JOB OLDjtl
Mfk*
IX
**«»..
®_
FINAL TEST VISUAL
PASSED
82 OR NEWER O
FAIIED
© CATALYTIC CONVERTER ©
© FUEL RESTRICTOR ©
© AIR SYSTEM ©
O 81 OR OLDER
©
FINAL TEST EMISSIONS LEVELS
2500 RPM*
SCO
©©•©
©©•©
©O
©•©
©•©
©•©
©•©
•81 OR
NEWER
Conpir*
tbtst
Imli
to lli>
enitsions
HindiTdi
lor modil
rait
b.ing
lutid.
ppmHC
©@®®
©©©©
©©©©
©®®
®®©
©©©
©©©
© PASS ©
"© FAIL ©HC
SCO
@©
-------
COLORADO "AIR" PROGRAM REPORT
NCS Trans-Optic B10-32551-321
, WTATWi
•*«**«
rr
WAT O
fOBD O
HONO 0
I»T£ O
JAGU O
jta> O
LAW O
UN* O
MAZ»Q
O
O
O
m«O
P58S O
PUGT
SUBAO
W O
TOP O
VDtK O
voiv O
mm Q
I 1 111
52 Wt *l8**ft 0
©« SYSTIM
oa mm
fJBST TtST t»ISSWNS IfVttS
llxie
-• nun,
LU
»*
Ul
*•*»
httli
urn
wiswwts
<$®
PASS
rail
IMPORTANT - KEEP THIS FORM
It will allow you to have ona free re-
inspection at the original inspection
station within ten days if your vehicle
fails its first inspection.
©_
FINAL TEST VISUAL
82 OR NEWER O
PASSED FAILED
© CATALYTIC CONVERTER ©
© FUEL RESTRICTOR ©
© AIR SYSTEM ©
OBI OR OLDER
FINAL TEST EMISSIONS LEVELS
2500 RPM-
•/• CO
©©•©
®Q0>
@©
©•©
0®
©•©
©•©
©•©
•81 OR
NEWER
Compart
thin
iBVtlS
loth.
emission!
slindiids
lor model
re«r
being
tested.
iHC
©©©©
00O©
©@©©
©@©
©0®
©@©
0©0
®@©
©@©
CO
© PASS ©
© FAIL ©
HC
V, CO
©©•©
oo-o
©•©
©•©
©•©
©.©
©•©
©•©
IDLE
2nd'
Compare
hese
rts
othi
missions
tanderds
rodel
being
tested.
ppm HC
©©©©
©O©Q
©©©©
©@©
©©©
©@©
0©0
CO
© PASS ©
© FAIL ©
HC
© "AIR" PROGRAM ADJUSTMENTS
STATION
NUMBER
MECHANIC'S
NUMBER
ADJUSTMENT
COST
s .
s |*
«
© EMISSIONS
REPAIR COST
(81 OR NEWER)
NOTt TO VEHICLE OWNER
II your vehicle did not peas the lirst test, you may make eny adjustments or repairs yourself
or have anyone else do them but. in order to receive a Certification ol Emissions Control.
your vehicle must meet at least one ot the following conditions:
2) Pass the visual final test and be adjusted by a licensed emissions mechanic
and. lor 19B1 and newer vehicles, have at least one hundred dollers {S100I
in emissions-related repairs done.
fOR ANY ADDITIONAL INFORMATION. CONTACT DUE OF THE FOLLOWING EMISSION
TECHNICAL CENTERS:
Denver Metro Area - 9640 C. Collax Ave.. Aurora 364-4135
Fort Collins Area • 429 N College Ave.. Ft. Collins 221-5324
Colo Springs Aria • 1403 S. Tejon St.. Colo. Springs 631-2333
DR THE COLORADO DEPARTMENT OF REVENUE 866 551B
EMISSIONS STANDARDS FOR THIS VEHICLE
HC .
CO
with ftf ralii **4 gtilfoltMi of tin CeJonrf* *18 tt«f>*«>-
SrSJWTUR! « WCUtSEft ifWttSM* ««tlAHlC
THIS COPY TO BE RETAINED BY VEHICLE OWNER
158
PAGE 3
-------
INSTRUCTIONS TO LICENSED EMISSIONS MECHANICS
NOTICE: Only a Licensed Emissions Mechanic may perform this inspection.
• Page 1 (top copy) of this form will be scanned electronically.
• Please keep page 1 free of dirt or grease.
• Do NOT fold, staple, spindle, or mutilate page 1.
• Do NOT make ANY marks on page 1 other than specified in
directions.
• You may make notes on pages 2 and 3 ONLY.
• Use ONLY a NUMBER TWO pencil on this form.
Where boxes are provided put one and only one number or letter
in each box, if there are more boxes than numbers or letters put
enough zeroes in front (to the left) to fill up the extra boxes. For
example, if the emission levels are 4.8% CO and 750 ppm HC.
ENTER
DO NOT ENTER
SCO
ppm HC
07
% CO
4 8: o
ppmHC
After the boxes are marked, fill in the circle under each box which has a letter or number that matches the letter or number in the box above. There
should be one and only one circle filled in under each box.
FIRST INSPECTION
PRINT - Name and address of vehicle owner at top of this form.
ITEM
Q Vehicle Identification Number • Write in VIN number.
© Auto Make • Fill in circle for auto make.
0 License Plate Enter license plate number and fill in circles (if
temporary tag leave blank, if more than six characters use first six).
0 Dale of Test - Enter month, day and year and fill in circles.
© Station Number • Enter your station's license number and fill in circles.
© Mechanic's Number • Enter your emissions mechanic's license number
and fill in circles.
0 Model Year • Enter last two digits of vehicle's model year and fill in
circles (example: 1972 enter 72).
0 No. of Cyl. • Enter number of cylinders in engine and fill in circles (if
rotary engine enter number of rotors).
0 Inspection Cost • Enter cost of inspection in dollars and cents
(maximum $10.00) and fill in circles under dollar portion only.
© First Test Visual •
1981 AND OLDER VEHICLES: Fill in "81 OR OLDER" circle. (No visual
inspection needed.)
1982 AND NEWER VEHICLES: Fill in "82 OR NEWER" circle, complete
first visual inspection and fill in a pass or fail circle for each
emissions control system.
© First Test Emissions Levels -
1968-1980 VEHICLES: Complete first test emissions inspection, enter
readings in 1st IDLE boxes and fill in proper circles below.
Compare readings with State emissions standards and fill in pass
or fail circles for CO and for HC.
1981 AND NEWER VEHICLES: Complete first test two speed emissions
inspection, enter readings in 1st IDLE. 2500 RPM and 2nd IDLE
boxes. Fill in the proper circles under 2500 RPM • %CO and ppm
HC (left half of block ©). From the right half of block (n)
select the lowest IDLE CO reading of the two idle tests and fill in
the proper circles under IDLE %CO. Then select the lowest IDLE
HC reading (from either idle test) and fill in the proper circles
under IDLE ppm HC. Compare 2500 RPM and IDLE readings
marked in the circles with State emissions standards and fill in
CO and HC pass or fail circles for each speed. (Vehicles must
pass at both 2500 RPM and idle to pass the first test.)
NOTE: THIS COMPLETES FIRST INSPECTION. IF VEHICLE HAS PASSED
ALL REQUIREMENTS. FILL IN "COMPLIANCE" CIRCLE IN ITEM ® AND
SIGN ITEM © ENTER CERTIFICATE OF EMISSIONS CONTROL STICKER
NUMBER AT TOP OF FORM AND AFFIX STICKER TO THE WINDSHIELD
AS SPECIFIED. GIVE PAGE 3 OF REPORT FORM TO VEHICLE OWNER.
VEHICLES FAILING FIRST INSPECTION
ITEM
© "AIR" Program Adjustments • If adjustments were made by a licensed
emissions mechanic, enter license numbers of station and mechanic
where adjustments were made and cost of adjustments in dollars and
cents (maximum $15.00). Fill in circles under station number, mechanic's
number and dollar portion of adjustment cost.
NOTE: If adjustments were made by another licensed emissions
mechanic, copy this information from "Adjustment Verification Form".
© Emissions Repair Cost • For 1981 and newer vehicles only, enter cost
of emissions related repairs in dollars and cents and fill in circles under
dollar portion only. Repair costs do not include inspection or initial
adjustment fees, or any costs required to pass the visual inspection.
© Final Test Visual •
1981 AND OLDER VEHICLES: Fill in "81 OR OLDER" circle. (No visual
inspection needed.)
1982 AND NEWER VEHICLES: Fill in "82 OR NEWER" circle, complete
final visual inspection and fill in a pass or fail circle for each
emissions control system. (All three systems must be marked
"PASSED" in order for a certificate of emissions control to be
issued.)
© Final Test Emissions Levels - If adjustments were made by another
licensed emissions mechanic, copy data for blocks © and © from
"Adjustment Verification Form."
1968-1980 VEHICLES: Complete final test emissions inspection, enter
readings in 1st IDLE boxes and fill in proper circles below.
Compare readings with State emissions standards and fill in pass
or fail circles for CO and for HC.
1981 AND NEWER VEHICLES: Complete final test two speed emissions
inspection and enter readings and fill in circles as described in
item © . Compare readings with State emissions standards and
fill in CO and HC pass or fail circles for each speed.
© Certificate Issued - Fill in circle next to:
COMPLIANCE Only if item © passed and item © is within
State emissions standards.
ADJUSTMENT Only if item © passed, item © failed, adjust-
ments made by a licensed emissions mechanic and,
for 1981 and newer vehicles, at least $100 was
spent on emissions related repairs.
DENIED Only if item © failed or item © failed and 1)
adjustments not made by a licensed emissions
mechanic or 2) for 1981 and newer vehicles, less
than $100 was spent on emissions repairs.
© Signature of Licensed Emissions Mechanic • After completing this form.
sign it, enter number of the sticker issued (or "none" if certification
denied) at top of report form and give page 3 to vehicle owner. If a
certificate is issued, affix certificate of emissions control sticker to the
windshield as specified.
159
-------
APPENDIX E
161
-------
The following excerpts are from Forms Design Guide: Reference
Manual, Sentry Optical Mark Reading Systems, a traininq booklet deve-
loped by National Computer Systems, Inc. (NCS) for its customers. It
provides guidelines for designing forms for the NCS optical mark
reading system and in doing so, provides information about optical
mark systems in general.
Reproduced with the permission of National Computer Systems, Inc.
The interested reader is advised to obtain the entire document from NCS.
163
-------
-------
OVERVIEW OF SCANNING
WHAT IS SCANNING?
Scanning is the process of reading information from a document using either an optical mark
reader or an optical character reader. This process involves changing the information on the
document into a set of electronic signals which can then be stored in the computer's memory.
Simply stated then, scanning is another word for inputting data into a computer.
Scanning however, has a number of advantages over other standard means of input, such as
keypunching. Because the answers marked by the respondent are read directly by the scanner,
the need to transcribe data is eliminated. Obviously the data will get into the system faster and
with fewer errors. The net result is faster turn-around time and more accurate data. Because
scanning requires only a paper and pencil for input, the costs of data collection are reduced.
The primary purpose or the objective of scanning is to translate data on a sheet of paper into
information a computer can understand. In order to accomplish this, a certain sequence of
events must take place.
The first step in the scanning process is to design a form that will collect the desired information.
These documents may be such forms as test answer sheets, order entry forms, or staff surveys.
Secondly, once the form is completed, the programmer writes and tests a program to read these
documents. After this stage, the documents themselves are filled out.
The fourth step is the actual scanning itself. During this step the documents are processed and
an output record is obtained. This record consists of all the information you desired from the
forms.
The final step is the processing of the information by your main computer. This processing will
format the information into usable data. It is at this stage that your complete written report of
results will be produced.
WHAT IS A SCANNER?
A scanner is a system designed to read marks on a sheet of paper. This system usually consists
of several parts: A document feed mechanism (input hopper), document transport bed, scan-
ning station (read head), document output trays (output hoppers), and some sort of controlling
logic. The document feed starts the forms into the scanner, the transport bed carries the forms
through the scanner, and the output trays receive the scanned forms.
NCS scanners are called optical mark readers (OMR). This is in contrast to alternative readers
known as optical character readers (OCR). An OMR is a device that detects the absence or
presence of a mark, but not the shape of the mark. An OCR is a device that resolves the actual
character printed.
The scanning station or read head consists of a collection of light sources called light emitting
diodes (LEDs) and a collection of photo-sensitive cells that convert the emitted light into electri-
cal signals. There are two basic types of reading techniques: reflected light read and transmitted
light read.
2-1
167
-------
PAPER
' (CROSS
• SECTION)
UNIFORM LIGHT
TRANSMISSION
UNIFORM LIGHT
REFLECTANCE
Figure 2.2 Reading Techniques
In the reflected light reading system, there are two lamps placed above and at 45° angles to the
document. These lamps serve as the light source. Between these lamps is a photocell which
serves as receiving station. As the "white" part of a document passes under the photocell, a
certain amount of light is reflected into the photocell from the lamps. When a "mark" on the
document passes under the photocell, the amount of light reflected into the photocell is quite
a bit less than that reflected from the "white" paper.
The same basic equipment is used in the transmitted light read as in the reflected light read with
the difference being its placement. In the transmitted light system, the lamps are above and
perpendicular to the document while the photocells are placed below the document. All Sentry
systems are transmitted light or Trans-Optic® OMR systems.
In the NCS system there are 48 lamps and 48 photocells. This means that there may be a maxi-
mum of 48 positions that can be read across the width of the document. One of these positions
is occupied with a timing mark that cues the scanner to read across the form. Therefore, there
are 47 possible positions for responses across the form.
16 LEVEL READ
As the document passes through the read head, a certain amount of light is blocked by the docu-
ment. This is called the paper level. As a "mark" passes through the read head, more light is
blocked. The amount of light blocked in each position is measured by one of sixteen values, 0,1,
2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F. The value 7 indicates a very light mark and F represents a
very dark carbon mark. This is the NCS sixteen level read which permits careful discrimination
between marks.
Noise (erasures, smudges, etc.)
Light Marks
Dark Marks
23456789ABCDEF
Figure 2.3 Mark Discrimination Levels
2-2
168
-------
INTRODUCTION
This manual is intended to provide the necessary information and specifications for the design.
of forms for NCS optical mark reading systems. It was designed to provide the information,
rules, recommendations, and mock-up techniques needed to design a form to meet the needs of
a given application.
The first three chapters of the manual introduce the user to the scanning process, an overview
of forms design, and the design process. These chapters are not intended to be a complete course
in the process of forms design. For instruction in techniques and demonstration of forms de-
sign, see the Forms Design Self Instruction Manual (202 147 906).
The remainder of the manual is organized into a reference manual format that will allow quick
location of specifications dealing with any given topic. Once again, it is assumed that the user
has had experience or instruction in the basics of forms design and will simply be referring to
specific recommendations and regulations for a particular application. Use of the table of con-
tents and index will facilitate finding needed information. With these guidelines, the designer
can then be free to allow ideas and needs of the application to structure the form.
1-1
169
-------
INFORMATION FROM USER
FORMS DESIGNER
MAIN COMPUTER
REPORT
Figure 2.1 Overview of the Scanning Process
1-2
170
-------
LEADING EDGE
SKUNK MARK
TIMING MARK
BIAS BAR
O
w
UJ
Q
D
CD
? • * ? I t i ? ? I ;
oooooooooooo
000©
©®0©©©©©0@
©0©©©©©©0©
©©©©©©©©©©
©©©©©©©©©©
©©©©©©©©©©
©©©©©©©00®
©©©©©©©©©©
©©©©©©©©©©
©©0©©©©©0©
©©©©©©©©©©
©©0©©©0©0©
©©©©©©©©©©
©©©©©©©©©©
©©©©©©©©©©
©©©©©©©©0®
©©©©©©©00©
©©©©©©©©0©
0©©©©0©©©@
©©©©©©©©0(3
®©®®©©©®®®®©®®©®00®©®®©®®®O
®@®®©©©®®®®®®®©®©©©®®®®©®©o
®©®®©©@®®®®®®®©®©0®®®®®©®®O
®®®®©©©®®®®©®®©®©0®®©®®®®©O
®©®®®©©®®®®@®®(9®00®®®®©®®®O
®©®®©©©®®®©@®®©®©0®©®®®©®®O
®©®®©©©®®®©®®®©®©0®®®®®®®©O
®®®®©©©®®®®®®®©®©0®©©®®©®®O
®©®®©©©@®®®@®®©®Q0®®©®©©®®O
®©®®©©@@®®®©®®©®©©®©©®®©®®o
®©®®©©©®®®®@®®©®00®®®®©®®©O
®©®®®©@©®®®@®®©®©0®®®®®®®©O
®©®®©©0®®®®@®®©®©0®®®®©©®©O
ffiU1mmmu1CJ1U1U'
CO~JO>Cn&GJN>~*
0> ©J> Q> ©>•©> ©> 0> ©> 0> ©»
©01 ©CD ©CD ©0) ©D ©D ©O ©0) ©0) ©OS
©o ©o ©o ©o ©o ©o ©o ©r> ©o ©o
©o ©o ©o ©o ©o ©o ©o ©o ©o ©o
©m ©m ©m ©m ©m ©m ©m ©m ©m ©m
3SS3S3S3S2
Q> ©> ©> 0* ©> 0> ©> 0> ©> ©>
©03 ©01 ©D ©CD ©01 ©03 ©01 ©OS ©OB ©00
©O ©O ©O ©0 ©O ©O ©O ©O ©O ©O
©O ©O ©O ©O ©O ©O ©D ©O ©O ©O
©m ©m ©m ®m 0m ©m 0m 0m ©m ©m
Q> 0» ©> Qj> ©» ©> ©> Q> 0> ©>
©os ©o> ©os ©oo ©os ©oo ©os ©oo ©a ©o
©o ©o ®o ©o ©o ©o ©o ©o ©o ®o
©o ©o ©o ©o ©o ©o ©o ©o ©o ©o
©m ©m ©m ©m ©m ©m ©m ©m ©m ©m
©> 0> ©» 0> ©> Q> ©> ©> 0> 0>
©OS ©OS 003 ©OS ©OS ©D ©OS ©01 ©OS ©OS
©O 00 ©O ©O ©O ©O ©O 00 ©O ©O
©o ©o ©o ©o ©o ©o ©o ©o ©o ©o
0m 0m ©m ©m ©m ©m ©m ©m ©m ©m
- g
.
0> ©> 0> ©»• Q> 0> 0> Q> ©> ©>
©OS ©OS ©OS ©03 ©01 ©OS ©OS ©OS ©o ©OS
©O ©O 00 ©O ©O 00 ©O ©O ©O ©0
©O ©O ©O ©O ©O ©O ©O ©O ©O ©O
©m ©m ©m ©m ©m 0m ©m ©m ©m ©m
0> ©> ©» 0> ©> ©J> 0> ©» ©> ©>
©D ©OS ©OS ©OS ©OS ©OS ©00 ©OS ©OB ©OB
©o ©o ©o ©o ©o ©o ®o ©o ©o ©o
©o ©o ©o ©o ©o ©o ©o ©o ©o ©o
©m 0m ©m ©m ©m ©m ©m ©m ©m ©m
©> ©> 0> ©> ©> ©> ©> 0> ©> ©>
©OS ©OB ©00 ©03 ©00 ©OS ©00 ©OS ©03 ©OB
©o ©o ©o ©o ©o ©o ©o ©o ©o ©o
©o ©o ©o ©o ©o ©o ©o ©o ©o ©o
©m ©m ©m ©m ©m ©m ©m ©m ©m ©m
-
©> ©> ©> ©> ©» Q» ©J> ©)> ©>> 0*
©03 ©OS ©OS ©OS ©03 ©03 ©03 ©03 ©OB ©OS
©O 00©0©0©0®0 ©O ©O ©O ©O
©O ©O ©O ©O ©O ©O ©O ©O ©O ©O
©m ©m©m©m©m©m0m©m©m©m
©> ©» ©> ©> ©> ©> ©> 0> Qj> ©>
©OB ©CD ©00 ©OS ©00 ©03 ©OB ©CD ©OS ©OS
00 ©O ©O ©O ©O ©O ©O ©O ©O ©O
©o ©o ©o ©o ©o ©o ©o ©o ©o ©o
©m ©m ©m ®m ©m ©m ©m ©m ©m ©m
coco>hja>cji^G>rsj-'
0> ©> ©> ©> ©> 0> ©> 0> ©> ©>
003 ©00 ©OB ©OB ©OB ©03 ©03 ©00 ©OB ©CD
©O ©O ©O ©O ©O ©O ©O ©O ©O ©O
©O ©O ©O ©O ©O ©O ©O ©O ©O ©O
©m©m©m©m©m©m©m@m©m©m
TRAILING EDGE
Figure 2.4 Parts of a Form
m
x
» >
3 O
i m
I '
3 (A
I'
(A
m
m
2-3
171
-------
FORMS DESIGN OVERVIEW
Before we enter into a discussion of the forms design process, the quality of forms, and types of
forms, it will be helpful to define common terms relating to forms.
Parts of a Form
RESPONSE POSITION (also bubble or response circle)
The circles on the forms that indicate the positions which are capable of being read by a
single photocell in the scanner.
TIMING MARK
One of the short black rectangles on the edge of the form. They are printed in machine-
readable ink on front and back of the form. The timing mark triggers the scanner to read
the response position in that row.
TIMING TRACK
The series of timing marks running in a straight line along one of the long edges of the
document. To be scanned, there must be a timing mark for each row of response positions
on the page.
SKUNK MARKS
Black marks which occupy one or more response positions, they uniquely identify each
form. They are printed in machine-readable ink on the front and back of the form.
BIAS BAR
A bar of ink, the color in which the document was printed. The bar has no other marks.
It runs the width of the document perpendicular to the timing track and is printed on one
side only, either front or back. The bias bar is used to ensure that each photocell is read-
ing within specific tolerances.
LEADING EDGE
The leading edge is the edge of the form which passes first through the read head of the
scanner. The leading edge is parallel to the X-axis of the form.
GUIDE EDGE
The guide edge of the form is the edge with the timing track. This is the edge that slides
along the edge of the document transport bed and is critical to correct alignment of the
form in the scanner. The guide edge is parallel to the Y-axis of the form.
TRAILING EDGE
This is the edge that passes last under the read head.
2-4
172
-------
RESPONSE MASTER (also called response matrix)
All the possible response positions on a single page. Across the form, perpendicular to
the timing track, the bubbles are spaced six per inch to exactly match the photocell spac-
ing in the scanner read head. Down the length of the form, bubbles and timing marks are
spaced at either six per inch to yield a uniform matrix or at five per inch so that this di-
mension is line printer compatible. For example, an 8 1/2" x 11" form is 47 positions
wide by 63 positions long for a total of 2961 possible response positions when using the
6x6 master.
Therefore, there are two basic masters or matrixes: 6x6 and 6x5. Only one split is al-
lowed, but may be placed anywhere along the timing track. The split master layout sheet
can be made by splicing one of each kind together. Spacing perpendicular to the timing
track must always be six per inch to match the read head spacing.
LAYOUT SHEET
A printed copy of the NCS response master. It shows all possible response positions and
timing marks. The response master is printed on both sides of the layout sheet for ease
of doing two-sided designs.
MOCK-UP
A pencil drawing on an NCS layout sheet that shows exactly where copy, lines, responses,
etc. are to be positioned.
2-5
173
-------
=ooooooooooooooooooooooooooooooooooooooooooooooo
I =00000000000000000000000000000000000000000000000
I =00000000000000000000000000000000000000000000000
K =00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
I =00000000000000000000000000000000000000000000000
I =00000000000000000000000000000000000000000000000
I =00000000000000000000000000000000000000000000000
I =00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
I =00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
-=00000000000000000000000000000000000000000000000
=00000000000000000000000000000000000000000000000
ooooooooooooooooooooooooooooooooooooooooooooooo
ooooooooooooooooooooooooooooooooooooooooooooooo
ooooooooooooooooooooooooooooooooooooooooooooooo
ooooooooooooooooooooooooooooooooooooooooooooooo
ooooooooooooooooooooooooooooooooooooooooooooooo
i; =00000000000000000000000000000000000000000000000
K =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
i: =00000000000000000000000000000000000000000000000
II =00000000000000000000000000000000000000000000000
K =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
K =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
i; =00000000000000000000000000000000000000000000000
i: =00000000000000000000000000000000000000000000000
K =00000000000000000000000000000000000000000000000
K =00000000000000000000000000000000000000000000000
K =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
K =00000000000000000000000000000000000000000000000
II =00000000000000000000000000000000000000000000000
I! =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
K =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
I! =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
i; =00000000000000000000000000000000000000000000000
1C =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
K. =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
K =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
i: =00000000000000000000000000000000000000000000000
IE =OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
ooooooooooooooooooooooooooooooooooooooooooooooo
ooooooooooooooooooooooooooooooooooooooooooooooo
ooooooooooooooooooooooooooooooooooooooooooooooo
Figure 2.5 Response Master
2-6
174
-------
SUCCESSFUL FORMS DESIGN
NCS scanners represent a "systems" approach to optical mark reading. This means that each
component of the system — scanner, software, and forms — must be designed to function well
if the system is to be used to its full potential.
The forms component is important because the forms are the means by which data is entered
into the scanner. The respondent, the person supplying the information, marks directly on the
form and the form is fed directly into the scanner. Because of this direct route of data, it is essen-
tial that the form be designed so that the respondent can understand the instructions, and can
mark the form quickly, completely, and accurately. If the respondents are confused as to how
to mark their responses and where to mark them, it is unlikely that accurate data collection will
be possible. While the scanner and software are designed to catch problems and identify errors,
no amount of system sophistication can be built to correct the inaccuracies and missing data
that results from a confused respondent.
The first requirement of a forms design then, must be that it clearly identifies its own function
(has a title), tells the respondent what is wanted (has instructions), and provides a place for a
response. A good or successful forms design must do more than that, however. It must also be:
-.Compatible with the scanner model being used
-Able to be programmed with efficiency
-Adapted to the age and environment of the respondents
-Adapted to special needs such as mailing or carbon copies
The above list is only a sample of the requirements that may be placed on a forms design. Further
discussion of this is found in the section on background information.
The NCS OMR System has been designed to give the forms designer as much flexibility as pos-
sible. While it is true that there are some basic rules that must be followed, the rules do not usu-
ally dictate design or how the form is used. The forms design follows functional requirements
that are defined for the application. A good forms design is one that does the required job of
data collection.
Example:
1. If line printing is required on a form, that function requires the form to be continuous
and the design to be compatible with line printer spacing. Once this requirement has
been met, the designer has many options as to the layout of the form.
2. If a name and identification number are data that must be collected on the form, there
are some standard ways to capture alphabetic and numeric data, but the designer is
free to place or to orient the data field in almost any way desired.
THE DESIGN PROCESS
There are several important steps to be followed in the process of designing a successful form:
I. Find out the background information. Who is the user? Who is the respondent? What
is the purpose of the form? How will it be used?
2. Identify some of the general form requirements such as size, content, and type of form.
3. Identify specific details of the design. Who will sign off on the design? What forms
handling equipment will be used? What scanner model? What scanner software?
4. Prepare a rough draft.
5. Prepare a second draft.
6. Prepare a final draft.
7. Review requirements against the final design.
8. Submit design to NCS for composition and printing.
-------
NCS is interested in providing the best possible forms composition and printing. To that end
we would like to ask you, the designer, to provide us with clearly drawn mock-ups and complete
specifications. When your mock-up is received at NCS, it is examined by experienced NCS
forms designers. If your mock-ups and specifications are clear and precise, they will go directly
to our composition department. If not, we will contact you with our questions.
Before the actual printing of the form, NCS will send you a copy of the final composition of the
form (called a proof) for your approval. Actual printing will not begin until we receive your
written approval on the proof.
FORMS QUALITY
There are NCS technicians at each stage of production to ensure that you achieve a high quality
form. In addition to helping you in the design process, they monitor the quality of materials
and printing.
Sentry scannable documents are printed on Trans-Optic® bond paper which is manufactured
exclusively for National Computer Systems. Since the Sentry scanners read through the paper,
it is important that translucency of the paper, as well as all other facets that affect the scanner
both electronically and mechanically, be carefully monitored.
the consistent quality of Trans-Optic® bond paper is ensured beginning at the paper mill where
the paper is produced. NCS has placed a unique transmitted light scanning device there to
monitor paper inconsistencies caused by inherent differences in the various pulps used and vari-
ations created in the mixing process. This checking ensures uniform readability of the paper on
which your form is printed.
The ink is monitored in a similar program. All ink is formulated for NCS by the ink manufac-
turer to meet scanner specifications. Since black ink must be read and-color ink must be virtual-
ly transparent to the scanner, each ink formula and batch is tested on Trans-Optic® bond and a
Sentry scanner to ensure that it conforms to specifications.
Finally, the paper and the ink are tested together on the finished form product. In conjunction
with this procedure, forms are pulled at regular intervals during printing and checked for ap-
pearance, registration, and read-head alignment.
NCS Scannable Documents Guarantee:
All scannable documents produced by NCS are subject to the highest stan-
dards of quality control. Every phase of production from the manufacture
of paper to the method of packaging is checked to ensure that all specifi-
cations are met. Check sheets from every printing are scanned prior to ship-
ment as a final quality control measure.
National Computer Systems has an outstanding reputation for service and
quality. In the unlikely event that a problem should occur, NCS will either
replace any unacceptable or unscannable documents or assume the costs of
alternative data capture up to a mutually agreed upon dollar amount.
2-8
176
-------
BASIC DESIGN PROCESS
The steps to be followed in the process of designing a successful form are described below.
BACKGROUND INFORMATION
Before a pencil ever touches the layout sheet, the forms designer must learn as much as possible
about the form and how it will be used. The more that is known about the requirements of the
application, the greater the chance that the design will be successful. Below is a list of some of
the most common points to investigate. As you develop your own applications, you will surely
add to the list.
GENERAL BACKGROUND
1. Identify purpose of form. What is the user trying to accomplish with the form? What
information does the user need?
2. Identify the user. Who will be using the information collected on the OMR form?
3. Identify the respondent. Who will be marking the form? What characteristics, such
as age and education, do the respondents have that may affect forms design?
4. Identify how the form will be used. What environment will it be marked in (class-
room, truck, cab, office, etc.)? How may the respondent be expected to handle it?
Will it also be handled by someone other than the respondent? Will it be mailed?
Does the respondent have previous experience with OMR forms?
5. Identify other overall requirements. Are there cost limitations? When are finished
forms needed?
SPECIFIC ELEMENTS TO BE INCLUDED:
1. What specific information is being requested on the form?
2. What format is required for each piece of data?
3. What instructions will the respondent need to complete the form correctly?
4. What marking instructions and examples will be needed?
5. What type of form is required? Cut? Continuous?
6. What size is required? Does size meet scanner and printing specifications?
7. Is there a desired sequence for data fields on the form?
8. Will all text be included on the form or is this to be an answer sheet with text on a
separate page?
9. Are the specifications flexible? How much freedom does the designer have?
OTHER DETAILS
1. The forms designer will need to consult with other people involved with the form,
such as the Sentry Programmer. Identify all such people and make sure they have ap-
proval responsibility over the final design.
2. How will the form be handled? Will it be mailed? Folded? Line printed?
3. What model scanner will be used? How will scanner specifications affect forms design?
4. What scanner software (programs) will be used? How will software specifications
affect forms design?
3-1
177
-------
ROUGH DRAFT
The purpose of a rough draft is to determine approximately how much space the response areas
will require, how they will fit on the page in general, and what space will remain for other need-
ed information. Each of the items on a form are sketched onto an NCS Layout Sheet.
Layout Sheet
To start a rough draft, begin with the NCS layout sheet. Your first question will be which master
to use, 6x6, 6x5, or split? The choice of master depends almost entirely on the requirements of
your design.
FORMS DESIGN HINT: Choice of layout masters
A 6 x 6 master has the greatest response density. It is used on most cut
sheets and allows the greatest amount of data to be on the sheet.
A 6 x 5 master is designed for line printer compatibility when it is oriented
horizontally. It may also be used for cut sheet forms where more space
between columns of circles is desired.
Split masters make the most of both formats: A 6 x 6 section for best use
of available space and a 6 x 5 section for line printer compatibility.
Fitting Form Parts
Since skunk marks and bias bar are critical parts of the design and suggested positions are shown
on the layout sheet, begin by marking them on the layout sheet. Next you fit in the grids. Fit-
ting the largest items first will enable you to see how much effect these areas have on the overall
design. Then the smaller grids can be sketched into appropriate locations.
You will find that, as you fit grids onto the layout sheet, each grid could just as well have been
positioned in some other location on the form, since there is no inherent restriction in location
or orientation. It is usually a good idea however, to have all areas of a form oriented in the same
direction so that the respondent does not have to turn the form from side to side.
Once you have your grids sketched, you must fill in the title and instructions. By rearranging
the grids for the best design, there should be adequate space for the title and instructions.
For your first designs, you may want to try several rough drafts for each form. Don't hesitate
to use as many layout sheets as necessary. They are provided by NCS at no charge.
FORMS DESIGN HINT: Cut and Paste
While many find this Rough Draft approach is the easiest way to work up
a design, others prefer the Cut and Paste method. With this method you
cut out pieces from an existing form that meet your new form requirements.
You then rearrange the pieces until you arrive at a pleasing and efficient
design. Finally, the pieces are pasted in place and your rough draft is almost
finished.
Even when the pieces are pasted to an NCS layout sheet, the pieces will not
quite actually fit the space. It is therefore necessary to take a fresh layout
sheet and draw an accurate mock-up. Doing this additional step will en-
sure that the design works the way you want.
Please do not submit cut and paste work to NCS as a final draft.
3-2
178
-------
NO-
j' f J*
WflME
s
c
o
o
u
R
T>
E
H
a
o
t
c
4
i
BIRT"
OKS %
MB I
mm I
INSTRUCTIONS
Figure 3.1 Sample Rough Draft
3-3
179
-------
SECOND DRAFT
Drawing the rough draft is only the first stage in the mock-up process. Returning to the original
draft, you must evaluate and make improvements on each section. It is at this stage that you
spread out some grids, squeeze others together, try different formats for grids, and make room
for additional information.
When the second draft of the form is complete, check .your progress:
Are all data elements included?
Is each element the correct size?
Are the elements in the desired sequence?
Are the grids spaced for easy readability?
Does the form fit the format specified?
Depending on your organization, you should seek approval from any individuals who must ap-
prove or sign off the design before the final draft is completed. For users who are not very
familiar with OMR forms, it may be necessary to complete a final draft before asking for their
approval.
ACTION
ST
-------
r 13 ir *? ** !• H
o oo oo oo oo oo
flCTIOtf
STUD£*T
No
FAMILY
NO
wm^m
I
C
«
C
o
L
N
o
•^
4-
a
A
P
^
'
12. II
OO
IS
O
Figure 3.3 Sample Second Draft
FINAL DRAFT
The second draft merely outlines grid locations. It contains none of the detail that will be re-
quired by NCS to typeset and compose the form successfully. The steps for the final draft are
as follows:
Step 1: First, on a fresh layout sheet, pencil in the exact outlines of the grids.
Step 2: The lines which will go inside each grid are drawn exactly where desired.
Step 3: Grid headings are added.
Step 4: Response circles and the response designators are placed. Some will have the
response choices listed with a circle beside each. Others will have some unique
characters to represent individual responses and these are shown exactly as de-
sired. Standard alpha or numeric responses may be indicated in one column and
then marked to indicate that they repeat. This type of grid is so common in OMR
forms that the NCS forms personnel will automatically provide a complete grid.
Step 5: The drawing is now complete. Other miscellaneous features may be added. For
example, if a corner cut is desired, that is marked on the mock-up.
Shading of alternate columns is desired to make it easier for the respondent to mark each column.
We suggest using a colored marking pen of the type used to highlight passages in textbooks or
notes. Do not mark the actual mock-up. Since the markers are usually indelible, it is a good idea
to make a photocopy of the original and then mark the copy. If shading changes are made, the
original mock-up is still usable. Various percentage values of shading and even reverse-type
areas can each be marked with different colors. Just make sure you provide a key to color
significance.
3-5
181
-------
FORMS DESIGN HINT: Text
When large amounts of text are required on the form, it would be quite
time consuming to write the text on the mock-up itself. In such cases, it
is more convenient to indicate the general layout of the type on the mock-
up and type the actual text on a separate page. Be sure to indicate where
the copy must be placed on the mock-up.
SEPARATE TEXT
MOCK-UP
*** <*<«•*
Figure 3.4 Text Indicated on Separate Sheet
3-6
.182
-------
STU6ENT
NtW O
TTUOCMT
WlHljtR
IITTT
I C
r
MhliV
TTTTT
C
UST NAHf
IIMIIIIII
2 " •' .'' • -''
c- •- ••• ,
1 :• •.. ..'• '. •.':
*
M ;
1 :' .
4' .'. •' '. -.
H;., '- . •' :;.•
fO ". ':> : 6
«... <"! '.. 6 ' '. • :•
To ' ' ' ''• ( ' •;' '!
ro ••:: o :: o ,•
M •:'; C •": •" ;
vc. 3 ,. c •:• )
wo c: o r c •
H •."'; ( ( ) ' >. ', C ' •
»•:• C r -c •'" o
FlRTT N«NB
Jlllllll,
°3-— T-*-?
' -..' ' *- S ' • *>'
.,., ., ,. .
"" "» '
f / " '• ,'
^ - o \>. ":•
V; ' ; '
.'! ('-" ) '....' ""'.1
' } :. ,' ; ^ ' ' ^!
C, •' ) . - : ;;
•••'."; O '.' O )
/-•• -, /••• >
./
••C 0 "^ '.' )
i C". • ~) ; i •' i
o -c :-• c o
• ',• . W"
LLJ
°r
MtQ
u
toy o
ytvo
NO o
DHC
MO. 2.
PENCIL
ONLY
10% SHADING
Figure 3.5 Sample Final Draft
3-7
183
-------
PROCEDURE CHECKLIST
When the mock-up is complete, the form should be checked against specifications a final time.
Also there is a checklist on the layout sheet to remind you of all typical forms specifications. This
checklist is not intended to be all encompassing; it covers only the most common specifications.
It will help you make your mock-up and specification complete and it will help NCS produce a
form that is exactly what you wish.
After users and other involved people have approved the final design, it is ready to be mailed to
NCS/ Data Forms Customer Service for composition and printing.
TRANS-OPTIC® FORMS DESIGN CHECKLIST
Before submitting your design to NCS, please make sure that each applica-
ble item on the checklist is clearly specified. Items which are not applicable
should be left blank. Items marked ® must also be marked on the mock-up.
Orders without a clear mock-up and complete instructions may be delayed.
Thank you.
OFORM TITLE '.—
OFORM SIZE (Including carrier strips if continuous.)
OFORM TVPF
OCut sheet (single page)
®Fold to ,
OContinuous (line-printer compatible)
®Pre-slug positions (. .. , . . ......
,*&.,, . Mndicate number of characters in each field.
QE>Pre-prmt positions)
(^Alignment notch
O Multi-part of parts.
mark one: O Snap-out or . O Continuous
mark one: O Carbon copies or QSelf contained copies
mark one: ®Spot carbon or OFull carbon
®Stub location - Indicate which side copies are to be attached.
Stub attachment (Mark one for each side)
O'eft side: O crimp Oglue Onot attached
Oright side: Ocrimp Oglue Onot attached
OCarbon length from stub end:.
O Caper and ink colors for carbon copies. Ink color choices are identical to
Trans-Optic" ink colors; paper color choices for copies are white, blue.
pink and canary.
Part 2 Ink: , Paper: •
Part 3 Ink: _ , Paper:
Part 4 Ink: .-, Paper:
Part 5 Ink: , Paper:
OTRANS-OPTIC"5INK COLOR: Front: Black and
Back: Black and
OTRANS-OPTIC®PAPER COLOR:
OWhite- Trans-Optic Bond* O Vio-Trans-Optic Bond*
ORESPONSE SPACING: Qe x 6 Oe x 5 Ospiit master
OSCANNING SYSTEM Form will be used with: IQ VALTREP System
OScanner model no ^ODOSSIER System
(OSheet Compile System
OMake sure that response positions, bias bar, skunk marks and write-in areas
do not back-up to similar areas on the reverse side of the form.
^NUMBERING
®Serial number from to
® Litho-code from to
®SKUNK MARKS
eBTIMMM^MABK*/*1-1- TIMING MARKS WILL BE PRINTED UNLESS
HCORNER CUT(sA°THERWISE MARKED ON THE MOCK-UP.
®CORNER TIP(S)
^DRILLING OR PUNCHING
®PERFORATIONS
® SHADING. A 10% value will be used unless otherwise noted.
O Proof should be sent to:
Name ,
Address.
O'f NCS has questions, who should we call?
Name
Phone 1
^Please indicate other specifications on the reverse side. If in doubt as to
any specifications, please call our Data Forms Customer Service at
800-328-6302. We will be happy to help.
Figure 3.6 Procedure Checklist
3-8
184
-------
PROOFREADING
The end product of NCS typesetting and composition is a proof, which is a blue and white photo-
graphic print of the completed form. The purpose of the proof is to show you exactly how the
form will look. It is your responsibility to check that the form is done exactly the way you want.
If there are minor errors (such as typographical errors) or changes to be made, make the correc-
tions directly on the proof. If the changes are at all major, such as rearranging grid positions or
rewording entire bodies of text, the changes should be typed out on a separate sheet or redrawn
on a fresh layout sheet.
Before returning the proof to NCS, you MUST sign it and indicate what category the proof is:
OK
OK with Corrections
Rejected
If you indicate OK, NCS will go ahead with the printing of the form. If you indicate OK with
Corrections, NCS will make the small corrections, send you a confirming proof, and print the
form. If you have indicated that the proof is Rejected (that is, it needs many major changes),
NCS will make the changes and send you a new proof for your approval.
Keep in mind that while NCS will always take responsibility for the basic scannability of the
form, you, the customer are responsible for the form design once you have signed your approval
on the proof.
CUSTOMER PROVIDES CAMERA-READY COPY
If desired, a customer may supply "camera-ready copy" for development purposes. Since there
are many defined conditions which must be met, contact NCS for specifications before starting
forms development.
3-9
185
-------
(INTENTIONALLY BLANK)
186
-------
TYPES OF FORMS
Optical Mark Reading has been applied to a tremendous number of data entry needs. OMR
forms applications include grade reporting, course selection, election ballots, vehicle registra-
tion, payroll deductions, retail inventory, service reports, and water meter reading, to name just
a few. Different applications require different types of forms. A grade reporting form will have
very different requirements than an election ballot. The NCS Applications Manual (NCS part
no. 202-147-443) contains examples and descriptions of more than 50 different applications.
We hope these examples will stimulate development of new applications and design ideas.
NCS scannable documents are manufactured in four basic formats: cut sheets, continuous,
saddle stitched booklets, and multi-part forms. Each format and its specifications are discussed
below.
CUT SHEETS
Cut sheets are usually cut to size during printing. They may be single sheets or folded.
Single Sheet Folded
Figure 4.1 Cut Sheet Forms
CONTINUOUS FORMS
Continuous forms are designed to be run through a computer line printer prior to being given to
the respondent. Some scanner-readable information, such as respondent identification, is print-
ed. The respondent then marks the form and returns it to the user to be scanned. For this reason,
such forms are sometimes called "turnaround documents".
r>!= •
o! =
o! =
oil
O' —
°;i
o.=
o! =
ol =
o!= .
o' ~
oi =
in
IO
So
!o
jo
!o
1°
10
!o
I
Figure 4.2 Continuous Forms
Continuous forms are manufactured from a continuous web of paper but are not cut into units.
Each form is separated by a perforation where it can be torn apart after it has been run through
a line printer.
In addition, continuous forms have a perforated strip on each side called "carrier strips" or "pin-
feed strips" that enable the line printer to transport the paper during preprinting and preslugging.
4-1
187
-------
CARRIER
STRIPS
PERFORATIONS
o; iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiniii'o
llllllillllliillllllillllllllllllilllli
Figure 4.3 Continuous Form Carrier Strips and Perforations
Format for Continuous Forms
Line printer spacing is typically 6 lines per inch vertically and 10 characters per inch horizontal-
ly (with 132 characters total), while forms are either 6 x 6 or 6 x 5. Therefore, the forms designer
must take particular care to ensure that each continuous form is line printer compatible, scan-
ner compatible, and printing press compatible.
Following are diagrams of some of the most common continuous form formats to show how
they are set up for line printer compatibility.
-------
814" x 12" - UM the 854" x 11" -6x6 layout sheet
12" « 17" - U<* the 11" > 17" - 6 x 6 layout than
- 6 par inch -
(Correspond! with
line printer and
read-head spacing)
o|
Ol
oj
Oi
oj
Ol
Oj
Oi
0 |l 1 1 1 1
!o
'o
!o
10
1°
10
S~ TIMING TRACK 1 °
r 10
1 1 1 II 1 1 1 1 II 1 1 l|o
• 5 per inch •
(Each response corresponds with
every other print position on line printer)
9%" x 11" - Use the 854" x 11" -6x6 layout sheet
- 6 per inch -
(Corresponds with
line printer
spacing)
oir
o'r
o|E
°i-
O (~ TIMING TRACK
0|=
0 =
Oir
O'Z
o!=
!o
jo
|0
IO
|0
10
!o
!°
. 0
IO
!b
• 6 per inch-
(Every 3rd response corresponds with
every 5th print position on line printer.
Also corresponds with read-head spacing.)
11" x 17-7/8" - UM the 11" x 17" -6x5 layout sheet
- 6 per Inch -
(Corresponds with
line printer and
read-head spacing)
Oi
°!
o!
o|
Ol
o!
°!
o!
o!
°!
oi r
0 !< 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
i°
!o
'0
|o
10
!°
!o
'0
!o
IO
TIMING TRACK 1 O
lO
III 1 II III! Ill II Illlo
- 6 per inch -
(Each response corresponds with
every other print position on line printer)
- 6 pit inch -
(Corresponds with
line printer
spacing)
ol-
oiE
0[E
°E
Ol-
O i
oiE
0 1^
Q J— . /- ""ING TRACK
O!E
ojz
OlZ
o '—
°iE
o
o'~
olE
to
jo
1°
1°
!°
IO
!°
o
1 w
IO
10
jo
|0
'0
|o
lO
!o
!0
(Evary 3rd response corresponds with
every 5th print position on line printer.
Also corresponds with reed-heed spacing.)
It is possible, on the 814" x 12" and the 11" x 17-7/8" documents, to have a portion of the
form 6x6 and the remainder 6x6. The important thing to reniambei here is that you must
remain 6 par inch in the direction of the read-head and line printer spacing.
EXAMPLE
- 6 per Inch -
oj
0"
o
01
°l
°!
0
0
I 1 1 1 1 1 1 1
!o
10
•is
10
!°
r TIMING TRACK | O
10
1 1 1 1 1 1 1 1 1 1 It 1 1 l|
6 par inch I 6 per inch -
NOTES: 1. The above approach must be limited to one splice.
2. Any 5 per inch (or 6 par inch) sections must run along the entire height.
3. When designing a non-standard continuous form, splice two appropriate mock-up
sheets into one before starting layout.
Figure 4.4 Standard Continuous Form Formats
Designing 12" x 17" Continuous Forms for Preslugging
12" x 17" continous forms may be arranged in two possible formats for preslugging. Design A
in Figure 4.S has the timing tracks at each 12" side. With this format, there is an extra space be-
tween the pages that throws off the 6 per inch spacing of the bubbles between page one and two.
This means that when the printer has preslugged page one and spaces down to page two, that it
will no longer be aligned with the bubbles on page two. One page may be preslugged and pre-
printed. The second may be preprinted only. >
Design B, however, has the timing marks for page one next to the center of the 12" x 17" sheet
and has eliminated the extra space. The line printer then can space directly down from page one
and preslug page two in one pass.
4-3
189
-------
B
•n
o'
1
01 *
•1
• j
o
0|
ol
1
o1
«!
0|
01
01
ol
0|*
«l 1
^. —
IMIIIIM Illlll !«
1°
l«
I*
lo
PAGE 1
0
1
0
1*
It
>0
«
PAGE 2 *\
i«
10
1 Mil 1 1 1 1 1 1 1 1 1 1 i»
1 9" — J*.
IN DESIGN B, BOTH PAGE
1 AND PAGE 2 MAY BE —
PRE-SLUGGED AT THE
SAME TIME
IN DESIGN A, EITHER
PAGE 1 OR PAGE 2 MAY BE
PRE-SLUGGED, NOT BOTH AT
ONCE
• 1
•1
•1
•l
1,
1
ol
ol"
0| •
1 1 1
o1
0
0
0
0
0
0|
01
Ol"
%l II
|o
|0
10
10
PAGE 1 lo
10
1 0
10
1 1 1 1 1 1 1 1 1 1 II 1 1 1 lo
0
*
0
0
PAGE 2 o
*
Q
It
1%
1 1 1 1 1 1 1 1 1 III 1 1 I 10
17'
12"
Figure 4.5 Preslugging 12" x 17" Forms
Designing 4-1/4" Height Continuous Forms for Preslugging
Continuous forms with 4-1/4" height present a unique problem. If they are printed at exactly
4-1/4", each form is scannable but only every other form is line printer compatible. Response
positions are spaced in increments of sixths from one form to the next and so are compatible
with the vertical spacing of the line printer.
To be properly aligned, alternate forms must be staggered in size, alternating 4-1/3", 4-1/6",
4-1/3", etc.
L_'l L1 JI L'J L'J i.'_' L1 _! 1 j. J
urn minium 11
COMPATIBLE
IIIIHMIIIMM
4 1/6"
4 1/4"
13
I
|3
H 11111IIIII III 111 i
4 1/3"
4 116"
41/4-1 ': LJHE-ifPiMrlSR
4 CQMPAimt
4 1/3"
n MM
o
0
0
*
0
o
a
0
0
3
3
3
3
C
3
,
HIM M M M 1 III! 1
HIM 1 1 M 1 1 Mill 1
HIM II M 1 1 Mill 1
HIM M III 1 MM) 1
III II M 1 1 1 1 1 II 1 1 1
3
3
3
"j
0
0
6
y
3
3
5
0
0
c
c
Figure 4.6 4-1/4" Height Continuous Forms
It is very important to understand that because of the uneven form height, these forms may not
be mechanically burst apart. They must be separated by hand.
4-4
190
-------
Indicating Line Printing Areas
To compose, and typeset line printer'.compatible forms properly, NCS must know how many
characters and which response, positions you intend to line print. The standard way of indicat-
ing this is to print a series of X's to show line print areas. Since your mock-up will be done in
pencil, it is helpful to mark each X in a'Contrasting color such as a red pencil.
WEEK
ENDING
MO.
X '
X
0 0
1
2
3
4
DAY
X
K
0 0
1
2
3
4
YR
X
K
0 0
1
2
3
4
DEPART*
.MKNT
NUMBER
X
0
1
2
3
4
X
«•
X
K
••» 0
EMPLOYEE
NUMBER
A
0
i
2
3
4
5
X
••M
x
•••
XI
.
<
— o1
i;
(
NCS WEEKLY TIME SHEET
SK LABOR
CODIS ON
REVERSE
•IOC
PRINT POSITIONS ARE MARKED WITH AN 'X'
0**
ONty
Figure 4.7 Line Printer Mock-up
4-5
191
-------
BOOKLETS
Scannable booklets provide a means to have very large amounts of machine-readable informa-
tion in a small package. They are particularly applicable in surveys or tests where it is desirable
to have text and responses on the same page.
Booklets are 11x17 inch cut sheets that are collated together, stapled, and then folded into book-
let form. Prior to processing, the pages must be cut apart at the fold so that each page is scanned
separately. Your NCS Forms Customer Service Representative can assist you in obtaining the
booklet cutting equipment necessary to do this. The critical guide edge of the form is always on
the outside. NCS can produce booklets from 4 pages (one 11" x 17" sheet) to 52 pages (thirteen
11" x 17" sheets).
SEPARATE 11 « 17" SHEETS
COLLATED TOGETHER
SADDLE STITCHED
FOLDED
Figure 4.8 Production of Booklets
4-6
192
-------
Booklet design is similar to other forms design. There are, however, some special considerations
you must be aware of before you do a booklet design.
1. The timing track and critical edge must always be on the outside edge because the spine
of the booklet will be cut off prior to scanner processing.
2. Orientation of skunk marks, bias bar, and timing marks should be the same for each
page of a booklet.
3. As with any form, each booklet page must have a unique skunk mark assigned to it.
It is sometimes helpful to assign the skunks in a systematic way that will identify indi-
vidual pages and individual booklets. The usual way to do this is to assign a series of
skunk mark positions that consistently indicate page sequence in the booklet, regard-
less of which booklet it happens to be. A separate series off skunk mark positions
would identify each booklet. These positions would be identical within a booklet, but
would vary from one booklet to the next.
BOOKLET
ONE
BOOKLET
TWO
• o • o o o o o o o o o o o o • • o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
• oo• o oooo ooo ooommo oo oo oo ooo oo ooooo ooooooo oooooo o
I
I
PAGE
IDENTIFIERS
I
BIAS BAR- - -
I
BOOKLET
IDENTIFIERS
I
SKUNK MARKS
Figure 4.9 Booklet Skunk Marks
4. It is important to have some kind of identification system to tie together the pages of a
booklet during scanning. In addition to skunk marks, there are two ways to handle
this:
a)Pinhole Numbers
Pinhole numbering involves punching a six digit number through all the pages of
the booklet. It is not machine-readable but does allow the pages to be visually gath-
ered together if they become separated. The largest booklet that can be pinhole
numbered is 24 pages (6 signatures). Pinholes can not be in areas with response
positions on either side of the paper.
4-7
193
-------
b)Write-in identification on each page
These are areas on each odd (or even) numbered page for the respondents to write
their names or ID numbers; again allowing scattered sheets to be visually gathered.
5. When processing booklets, it is often desirable to perform various edit checks, partic-
ularly on the accuracy of the identification information. To do this properly, the
booklet must be designed so that the identification page is the first to be scanned by
the system. This way, if some critical biographical information is incorrect, the entire
booklet can be selected out, either by a select stacker or by stopping the scanner.
Therefore, the identification page should be either the outside front cover or the out-
side back cover. Since skunk marks are usually placed at the leading edge of the form,
skunk marks are placed at the bottom of the page when identification is on the front
cover and at the top of the page when identification is on the back cover. This allows
the correct page to be fed into the scanner first. If the identification page is selected
out because of errors, the data pages which follow it may also be selected out.
6. Booklets of up to 24 pages may be pinhole numbered.
7. Booklets may be serial numbered or litho-code numbered but only on a single 11 x 17"
signature (4 pages). NOTE: Litho-Code requires a special set-up. If this is desired,
we must know before any development, not after the first proof. It is not possible for
NCS to print matching numbers on multiple signatures.
LEAD EDGE AT BOTTOM OF PAGE
LEAD EDGE AT TOP OF PAGE
Figure 4.10 Design of Booklet ID Page
4-8
194
-------
FORMS DESIGN HINT: Booklets
While booklet design is really not very different from other forms design,
it often seems more difficult because of the number of pages involved. Plan-
ning is the key to success. You must plan what will go on each page.
What do you do if your design just fits comfortably on 9 pages and you
would rather print an 8 page booklet than a 12 page booklet? Resist temp-
tation. Unless some copy can be removed or the design can be worked out
to fit 8 pages easily, you will be better off with the 12 pages. The respondents
will have an easier time marking the answers and the quality of the data you
get will show it.
Booklets are put together in increments of 4 pages (4, 8, 12, etc.) Often the
number of pages you require will not be an increment of 4. You may want
to use the extra space to spread out some areas that are a little cramped or
maybe you can put together some snappy graphics for the cover instead of
diving into the response areas right away. You can also leave blank pages
to fill out the booklet so that it is a multiple of 4. Don't let this bother you.
There is plenty of printed material that has a blank page or two.
MULTI-PART SETS
Multi-Part Sets are forms with copies and carbon attached. There are two types:
-Cut sets ("snap-sets")
-Continuous sets ("snapouts")
Cut sets (snap-sets) are usually made up of scannable top sheet and from one to five other parts
which are glued together at a stub attached to one side or end. Each copy is perforated at the
stub. Copies are made in two ways: Carbons can be interleaved between parts or the copies
themselves can be printed on a "self-acting" or "carbonless-copy" paper.
Continuous sets (snapouts) work in the same way, but have additional carrier strips for trans-
port through the line printer.
STUB
iiiiiiiiiiiiiiiiiiiiiiiiiiiiililiniiii
TRANS-OPTIC
(MUST BE TOP COPY)
CARBON PAPER
COPY
PERFORATION
Figure 4.11 Cut Set
4-9
195
-------
oi i n in i in mi i ii ii ii in ii inn inn 11111 mm 111 ii inn i n ii |o
o iiiiiiiiiiiiiiiiiiimiiiiiiimiiim
-CARRIER STRIPS
-TRANS-OPTIC
(MUST BE TOP COPY)
-PERFORATIONS
CARBON
COPY
Figure 4.12 Continuous Set
Multi-Part specifications are as follows:
1. Parts of the set are attached at a stub where they are glued for cut sets and either glued
or crimped for continuous sets.
If glued, only one side should be glued. Gluing is sometimes necessary if there is to be
extensive handling before detaching and scanning. Crimping is preferable when ease
of separation is desired, such as when forms are to be decollated after printing or on
one side of a continuous set with the other side glued so that it functions as a snap-out
after it is printed and burst apart.
2. Cut set stubs are usually 5/8" wide and may be on either the short or the long side of
the form. Continuous set stubs are 1/2" wide and correspond to the carrier strips.
One side of the continuous set is usually crimped for easier separation, especially if
the continuous forms are burst apart and then used as snap-sets after they have been
line printed.
THIS SIDE GLUED
iiniiiiiiiniiiniiiiiiniimmmiiinnnmiiiiiiMnii
o I nil n ill ill IIIIII n ill n ill M i n n inn ii ii il i ii i ii ii in mi
THIS SIDE CRIMPED
Figure 4.13 Multi-Part Continuous Set
4-10
196
-------
3. Copies can be made in two ways: Carbons can be interleaved between parts or the
copies themselves can be printed on self-contained or "NCR" (no carbon required)
type paper. NCS recommends using carbon paper because it is less expensive, more
available, and produces a sharper image than self-acting paper. Self-acting paper is
not recommended when there are more than 2 copies in a set.
4. When possible, the carbon paper will be left shorter than the original and copies so
that there is a grasping margin to allow easier separation.
5. Spot carbon is a means of deleting carbon from a specific area of a snap-set so that
marks made there do not mark through to the copies from the original. To indicate
where you want carbon, make a photo copy of your mock-up and mark the areas
which will copy and those that will not. See Figure 4.14.
_ NA H, «k. t>v U Urt. ^. •( 1
« ^A-\A.
ORIGINAL
Figure .4.14
COPY MARKED
TO SHOW CARBON
AREA
Indicating Spot Carbons
6. Maximum width of multi-part forms is 17 1/8", including the 5/8" stub. (Form is
16 1/2"). Maximum width of continuous sets is 17 5/8", including 1/2" carrier strips
each side. (Form is 16 5/8").
7. Copies are available on a 12 or 15 pound bond paper. Available paper colors are
white, blue, pink, and canary. Ink colors for the copies are identical to the choices
for OMR forms.
It is not advisable to use blue carbon with a blue form (body copy and response posi-
tions) because there is not enough contrast between the form and the marked response
position. This also applies when using black carbon with black type.
8. Marginal words may be printed on the form to identify copies such as "Data Proces-
sing Copy", "Office Copy", and "Student Copy".
9. Stub location is at the designer's discretion. However, we urge that it be located ad-
jacent to the trailing or outside edge to minimize problems caused by sets being im-
properly torn apart. The respondent may accidentally rip the skunk marks if the stub
is at the leading or guide edge.
4-11
197
-------
imiiimiiiiiiiiiiiiiiiiiiiiiiimm
iiiiiiiiiimiiiiiiiiiiiiiiiiiimmiil
Stub at Outside Edge Stub at Trailing Edge
Figure 4.15 Snap-Sets
NOTE: Because of the unique requirements of multiple part forms, it is always a good
idea to check with NCS prior to final forms design.
FORM SIZES
The sizes of forms may also vary. In deciding on a form size, you should consider what size can
be processed on your model of scanner (See Chapter 10) and what size can be printed by NCS.
Form sizes printed by NCS:
Height Minimum Width
3-2/3" 7-1/2"
4-1/4"** 7-1/2"
5-1/2" 7-1/2"
5-2/3" 7-1/2"
8-1/2" 3-2/3"
11" 3-2/3"
17" 3-2/3"
Maximum Width
17-7/8"*
17-7/8"*
17-7/8"*
17-7/8"*
17-7/8"*
17-7/8"*
17-7/8"*
Printed sizes always include stubs, carrier strips, receipts, etc.
* 17 5/8" if multi-part continuous sets, 17 1/8" if multi-part cut sets
** 4-1/4" height is impossible with continuous form. Individual forms alternate 4-1/3, 4-1/6,
4-1/3, etc. in height.
4-12
198
-------
T
CD
ui
1 T
.71/2".
IT
•
II
II 1 1 1 1 1 II 1 III
8'/2"
1
1
0|
0|
ol
o'
-1
0
°:"i
<0
!«
•
•«
ia
i°
III 1 1 II 1 1 III Mill |«
WIDTH
Figure 4.16 Width and Length of Typical Form Sizes
If a form is to be produced in a continuous or snap-set format, the scanning size of the form re-
mains the same as that of a cut sheet format. However, carrier strips or stub will have to be add-
ed to the overall width of the printed format. For continuous forms with a width of 7-1/2" up to
16-3/4",add 1/2" to each side. For forms wider than 16-3/4", add 7/16" to each side. Forsnap-
sets, a 5/8" stub is needed to glue the sheets into a set.
EXAMPLE:
8-1/2" x 11" in continuous format would be 8-1/2" x 12", or 9-1/2" x 11", depend-
ing upon carrier strip location. (Also applies for multi-part continuous.)
11" x 17" in a continuous format would be 11" x 17-7/8", or 12" x 17", depending
upon carrier strip location. (Also applies for multi-part continuous.)
8-1/2" x 11" in a snap-set format would be 8-1/2" x 11-5/8", or 9-1/8" x 11",
depending upon stub location.
4-13
199
-------
APPENDIX F
201
-------
The HCS price list reproduced on the following four pages is
provided only as an indication of what costs for optically read forms
might be. It is not definitive and is in no way binding to National
Computer Systems, Colorado State University, or the Environmental
Protection Agency.
203
-------
Trans-Optic
Custom Forms Price List
4401 WEST 76TH ST., MINNEAPOLIS, MN. 55435
612-830-7600
This price list covers the development,
printing and scheduling of all custom Trans-
Optic* forms (both new orders and reprints)
which are to be scanned on NCS equipment.
Your data form is a critical element in your
general scanning operation. The NCS com-
bined hardware and forms offering represent
the true systems answer to your scanning
needs.
Prices listed in this folder are presented
to assist you in determining your overall
costs and to help in planning your budget.
Prices may be subject to reduction based
upon the per order volume, forms sizes,
color combinations and delivery schedule
requirements.
EFFECTIVE JANUARY 1, 1982
Please direct written inquiries, purchase
orders and other materials to your local
Customer Service Representative. For imme-
diate assistance call Forms Customer Service
at:
800-328-6302 or
612-830-7610
800-233-0143 or
Minneapolis, MN
Lancaster, PA
Owatonna, MN
717-394-7196
800-533-0518 or
507-451-5137
We will respond quickly with a verbal
and/or written proposal. Thank you for the
opportunity to serve your needs and your
confidence in NCS and our products.
NOTE: Standardized Test Sheets and
General Purpose Answer Sheets
are carried in inventory and
listed on separate price sheets.
You may place an order or request
information by writing or calling NCS
Forms Customer Service.
NCS B-32263-321
Copyright. C NCS. Inc. 1981 ?05
All Right! R*i*rv>d
NATIONAL
COMPUTER
SYSTEMS
FORMS DIVISION
-------
PAGE 2
COMPOSITION/DEVELOPMENT:
•
MOCK-UP (LAYOUT)
COMPOSITION (INCLUDES ONE PROOF)..
i
' SHADING
i
1 ADDITIONAL PROOFS
REVERSE PRINTING
SPLIT MASTER
8'/2X
9'A X 1 1 or
ONE-SIDED
$ 75
$230
$ 25
$ 15
$ 25
$ 25
11
8V, X 12
TWO-SIDED
$125
$365
$ 35
$ 20
$ 35
$ 35
11 X
12 X 17 or
ONE-SIDED
$100
$335
$ 30
$ 20
$ 30
$ 35
17
11 X 17%
TWO-SIDED
$200
$590
$ 45
$ 30
$ 45
$ 45
FORMS REQUIRING ADDITIONAL COLOR WILL BE QUOTED UPON REQUEST (SEE "ADDITIONAL
COLORS" ON PAGE FOUR.).
ALTERATIONS - PRICE BASED UPON EXTENT OF ALTERATION.
ALL ART WORK. NEGATIVES, PLATES AND OTHER ITEMS PREPARED BY NCS SHALL REMAIN THE
PROPERTY OF NCS. ,
PRICES FOR OTHER SPECIAL DEVELOPMENT WORK WILL BE QUOTED UPON REQUEST.
ALL PRICES PER THOUSAND FORMS BASED UPON A MAXIMUM OF TWO COLORS PER
SIDE (ONE OF THE TWO COLORS ON EACH SIDE BEING BLACK)
QUANTITY
250,000
200,000
• 150,000
100,000
90.000
80.000
70.000
60,000
50,000
40,000
30,000
20,000
10,000
5,000
cm
8'/2 X 1 1
20.45
21.35
22.05
23.40
23.95
24.50
25.30
25.55
26.00
28.20
30.80
37.40
49.30
73.90
SHTH'S
11 X 17
36.05
37.80
39.80
43.25
44.85
46.50
48.10
49.65
51.10
54.95
57.50
61.50
75.20
108.00
9'/2 X 1 1
22.15
22.65
23.35
24.40
25.30
26.05
27.35
28.75
30.20
32.80
36.60
46.15
60.95
96.65
CONTINUOUS
8'/2X12 1
24.25
25.20
26.20
27.35
28.25
28.80
29.80
30,25
31.60
33.55
37.70
46.15
60.95
96.65
FORMS
1 X 177/8
35.20
36.40
37.90
39.35
40.55
41.55
43.65
44.85
46.45
49.00
50.90
61.50
90.15
142.10
12X 17
44.75
46.25
48.15
49.65
50.75
51.45
53.15
53.90
54.65
56.90
62.55
71.35
95.55
142.10
ALL PRICES ARE F.O.B. SHIPPING POINT UNLESS QUOTED OTHERWISE. TERMS: NET 30 DAYS
i CUT SHEETS:
, The above prices include plates, printing, paper.
I folding and standard packaging. Special features such
as numbering, special packaging, perforating, etc., will
be quoted on an individual form basis. Quantities and
sites not addressed above will be quoted upon request.
CUT SHEET STANDARD PACKAGING
QUANTITIES
Inner Outer
Forms Size Package Cartons
8-1/2 X 1 1 600 5,000
11X17 250 2.500
CONTINUOUS SHEETS: 1
The above prices include plates, printing, paper, perforating, line-hole punch-
. ing, and continuous folding. Special features other than those listed In this
paragraph will be quoted on an Individual basis. Quantities and sizes not
addressed above will be quoted upon request.
CONTINUOUS FORMS STANDARD PACKAGING
QUANTITIES
Outer
Forms Size Cartons
9-1/2X11 2.500
8-1 /? y 1 2 ? Rftft • '
11 X 17-7/8 1,500
12 X 17 1.500 or (1.250 if perl, at 8-1/2")
206
-------
PAGE 3
MOCK- UP - supplied by customer, or ....... ................................ ................... .Mlo w i 2 \veeks.
DEVO-OfiVlFN'' includes type, paste-up, negative assembly and proof ..... . ......... aiio . .; ^ we. ks.
(Major revisions to proofs may result in two additional weeks lead time for processing.)
PRiN i!MG subject to approval of proof or receipt of reprint order,
up to 250,000 ........ . ........................................ . ..................... uliow 2-3 weeks.
From 250,000 to 1 ,000,000 ........ .............. .................... ............. .illow 3-4 weeks.
Greater quantities quoted on individual basis.
Hifp!A;3 When placing a purchase order, consider the estimated ship date and your required in-
house date. Plan your method of transportation accordingly; i.e.: truck, parcel post, air
freight, etc. Specify method on your P.O.
. . ,• , Our overall lead time reflects the seasonal nature of your forms requirements. The quickest
turn-around time will be realized by placing orders from October through May. Orders placed
in the remaining five months may result in longer lead times.
If you need faster service, please* call for our best lead time estimate.
SPECIAL NOTES:
1. AVAILABLE PRICING DISCOUNTS
• You may also be eligible for a multiple order discount which is based on identical sizes and colors of forms.
The exact discount will be quoted by your Customer Service Representative at time of order.
• Our 4 for 4 program in 1981 — take advantage of a 4% discount on the printing price of your forms order.
To qualify, place your order at least four months prior to the requested ship date. This additional time
allows us to gain plant efficiencies which we will pass along to our customers: We will ship and invoice
upon completion of your order.
2. OTHER FORMS PRODUCTS
• This price sheet reflects only the most commonly purchased forms. NCS also manufactures booklets
(several sheets stapled together) and multi-part sheets (which include carbons or self contained reproduc-
tion paper). The multi-part documents are produced in either a continuous or snap set format. Due to the
specialized nature of these products they are not referenced individually. However, your Customer Service
Representative will be happy to provide pricing quotes and assist with placing your order. NCS intends to
meet all your scannable forms needs.
• NCS also manufactures standard one-part Green bar computer stock paper. Product specifications are as
follows:
Description Price per thousand
12" X 8-1/2" White Bond Green bar $6.32
14-7/8" X 11" White Bond Green bar $8.32
This product will be packaged 3000 per carton and will be shipped F.O.B. Owatonna, MN. There is a
minimum order quantity of two cartons.
3. INFORMATION REQUIRED WHEN ORDERING
• Form Size
— In the case of continuous forms, state overall sizes before and after carrier strips are removed.
• Color Desired
• Quantities Desired
— Higher volume purchases reduce your per form costs.
• Special Features
— Corner cuts, litho-code, perforations, alignment notch, etc., see back page for details.
• Include Copy
— Mock-up, previously printed form, etc.
• Ship Date Desired
— Plan to receive your forms several weeks prior to distribution and use.
• Receiving Location(s)
• Purchase Orders
— Should contain all standard or quoted prices.
REFER TO PAGE 4 FOR SPECIAL AND STANDARD FEATURES
-------
PAGL 4
FJEAjTUJRESj (Offered at an extra charge, please consult your Customer Service Repre
sentativR for details. Also, for a more complete definition of features, please refer to your Forms
Design Guide.)
• !•:'!lIQ^CpOiMG: A scanner readable numbering technique developed to identify two halves of a
sheet too large to scan on systems other than a model 7015. Litho-Coding can also be used
for special projects where machine readable identification numbers are required. Includes
matching decimal numbers.
• DEC I MAI. NU!ViBL:RUvjG. Available in vertical or horizontal format on most cut or continuous
forms.
• ^pKj^ii!:'.1 •-''•'}'~l-"'.("'V: This process is not required in the scanning operation but is often desired
by users to assure forms are stacked properly prior to processing.
• ALIGNiyttfv'i f'iOTCH: 1/4 inch holes positioned over the perforation that separates the form
from the carrier strip on continuous forms. The alignment notch serves the same purpose as
corner cutting but has the advantage of being placed in one of several locations along the car-
rier strip. This service is free on continuous forms but there is a charge on cut sheets.
• NOjM V>T..:' )•>;.; i-ACi;.A3IN«.'?: We welcome the opportunity to perform special handling, pack-
aging, and shipping requirements to fill your special need.
• .i>;..T!>~ . .;,: The standard prices on page 2 cover the printing of forms with a maxi-
mum of one color (plus black) on each side of the form. Extra colors are available and will be
quoted upon request.
STANDARD FEATURES:
» HIGHLY TRAINED AND DEDICATED PERSONNEL: The development and production of
Trans-Optic® forms requires very sophisticated graphic arts equipment. However, it takes
people, quality people, to monitor this equipment, arid we are proud of our training proce-
dures and employee dedication that result in the production of forms which are unequaled
in the industry.
0 QUALITY CONTROL PROCEDURES: We don't just talk quality, we produce it. Every phase
of production, from the manufacturing of the paper to the shipping of finished products, is
subjected to the highest standards of quality control. We see your forms the way your
scanner sees them.
« SCANNABLE PAPER: All NCS scannable forms are printed on NCS Trans-Optic® Bond
paper stock. This paper is manufactured exclusively for NCS and is monitored at the mill
using NCS testing equipment. Forms printed on non-NCS paper have caused scanning in-
accuracies and may result in unusable documents.
Our registered Trans-Optic® trademark on your forms is a guarantee that the paper opacity
and uniformity is within NCS system requirements.
" GUARANTEE: We guarantee our Trans-Optic® forms and back this guarantee with the
knowledge necessary to respond effectively when a problem is encountered. We have
highly skilled customer engineering personnel located throughout the United States who
are available to help when needed.
208
-------
APPENDIX G
209
-------
The following pages appear in the Colorado Department of Health
Mobile Sources Section report titled "Data Management for the AIR
PrograiTi, July 1981 to December 1981" dated January 20, 1982. They
describe the reports produced for the "blue system" for the AIR
Program and contain samples of several of the reports.
211
-------
AIR PROGRAM DATA REPORTS
}rogram Name
RIUBR1 AIR-Error Report
Due Dates Language Contents
Weekly COBOL Non-cumulative distribution by mechanic number within
station number of:
1) Tests added
2) Number of errors
3) Distribution of error percentages over:
a)make, b)vehicle license number, c) date,
d) cost, e) number of cylinders
4) Average first test CO and HC readings
5) First test pass rate
6) First visual fail rate
7) First emissions fail rate
8) First test CO fail rate
9) First test HC fail rate
10) Final test fail rate
11) Final emissions fail rate
ro
I—*
CO
RIUBR2 Statistics on Tests
by Station Number
Monthly
SPSS Cumulative station number distribution of:
1) Total tests
Total retests
Test fail rate; and for CO and HC
i
} Average first test readings
) Average final test readings
6) Percentage reductions
7) Actual reductions
RIUBR3 Statistics on Tests
by Model Year
Bi-monthly SPSS
Cumulative model year distribution of:
1) Total tests
2) Total retests
3) Test fail rate; and for CO and HC
4) Average first test readings
5) Average final test readings
6) Percentage reductions
7) Actual reductions
UUBR4 Pass/Fail Distribution Monthly
Frequencies
SPSS
l)Pass/fail counts for first and final test CO and HC
2)Frequency distributions for model year and CO and
HC and final tests and reductions
3)Count and failure rate for first and final tests
-------
AIR PROGRAM DATA REPORTS (continued)
Program Name Due Dates Language
RIUBR5 AIR-Monthly Error Summary Monthly SPSS
Contents
Monthly summary of errors, CO and HC readings, and
pass/fail rate distributions by mechanic number within
station number (weekly error reports) showing number
of errors in place of percentages and indicating
which values lie outside of acceptable tolerances.
RIUBR6 Distribution by Make
within Model Year
Monthly and
Quarterly
SPSS Cumulative and non-cumulative distribution by Make
within model year for CO and HC each of:
1) Average first test readings
2) Average final test readings
3) First test to final test reductions
4) Percentage reductions
ro
RIUBR7 Statistics on Tests
by County
Monthly
SPSS
Cumulative model year distribution by County of:
1) Total tests
2) Total retests
3) Test fail rate; and for CO and HC each
4) Average first test readings
>) Average final test readings
5)
I
Percentage reductions
7) Actual reductions
As requested
1) Average costs for: a)inspection, b) adjustment
2) Pass/fail rate for "home adjust" vehicles
3) Number of voluntary repairs
4) True (as opposed to issued) numbers for vehicles
a I Compliance
b) Adjustment
c) Denied
-------
SCHEDULE OF AIR PROGRAM DATA REPORTS
Day
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Program
iiiiiiiiiiiiiiiiiiiiiiniiiiii
iiiiiitmiiiiiiiiiiiiiiiniii
iiiiiiiiiiiiiiiiiiiiiiiiiiini
iiiiiiiiiiiiiiiiiiiimiiiiiii
iiiiiiiiiiiiiiiiiiiiiiiiiiuii
iiiiiiiiiiiiimmiiiiiiiiiiii
iiiiiiiiiiiiiiiiiiiiiiiiiiiini
iiiiiiiiiiiiiiiiiiiiiiiiiiini
IIIIIIIIIIIIIIIIIIIIIIIIIIUII
iiiiiiiiiiiiiiiiiiiiiiiiiiiin
!)uarterl y niiiiiiiiiiiiin
31-annual iniiiiniiiiiii
RIUBR1
lining
iiiiiiiiiiiiini
iiiiiiiiiiiinii
iniiiiA
iiiiiiiiiiiiini
iiiiiiiiiiiiinii
iiuiiAliiiiii
iinilA
iiiiiiiiiiiiin
llll(£k
iiiiiiiiiiiiiin
iiiiiiiiiiiiini
RIUBR2
iiiniiiiiiiiii
iiiiiiiiiiiiini
niiiiiiiiiiiiiii
iiiiiiniiiiiiiii
iiiiiiiiiiiinii
minimum!
iiiiiiiiiiiiini
iiiiiiiiiiiinii
RIIJBR3
Illllll^
iiiiiiiiiiiiini
iiiiiiiiiiiiini
iinijA
iiiiiiiiiiiiiin
mi. A
iimiiniiim
iniiiiiiiiiiii
RIUBR4
iiiiiiiiiiiinii
imtA
iiiiiiiiiiiinii
IIIIIHIIIIIIII
Illlllllllllllll
RIUBR5
IIMII^ HUH
iiiiiiiiiiiiiiii
iiiiiiiiiiiiiin
IIIIIIIIIIIIHII
RTIIRRfi
iiiiiiiiiiiiini
iiiiuA
iiuii^
iiiiiiiiiiiiiin
RTIIRR7
mii|^
iiiiiiiiiiiiiin
RIUBR8
iimiA
-------
R. IUBR1
REPORT DATE 1 1/16/81
STN » MECH *
1001 0741
STATION TOTAL
1003 0535
2566
2571
STATION TOTAL
1004 0964
1559
1562
STATION TOTAL
1006 1623
STATION TOTAL
£J 1007 2354
CT> 2356
STATION TOTAL
1006 2106
24 1 2
STATION TOTAL
1009 0666
c
; STATION TOTAL
a 1013 0316
0396
STATION TOTAL
1014 1523
1596
16S2
1 685
1 940
STATION TOTAL
1015 1916
1922
STATION. TOTAL.
RECS.
ADDED
1
1
5
2
1
8
2
1
2
5
1
1
2
1
3
1
1
2
2
2
5
1
6
1
1
4
2
1
9
1
1
2
REC.
ERRS.
1
1
1
1
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
0
3
1
b
0
0
0
1
i
0
1
C 0
LORADO DEPARTMENT
INSPECTION - MAINTENANCE WEEKLY
** ERROR PERCENTAGE BY FIELD »*»
MAKE LIC. DATE COST «CYL
O
0
0
0
0
0
0
O
0
O
O
O
0
0
0
0
O
0
O
O
0
0
0
0
O
0
0
0
O
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
0
0
0
0
11
100
0
50
100 0
10O 0
0 20
50 0
0 0
13 13
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
00
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 100
0 0
0 50
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20
0
17
0
0
0
0
0
0
0
0
0
OF HEAL
ERROR REPORT
AVG . COUNTS
X/CO X/HC
0.5 600
0.5 600
2 . 0 624
4.3 236
5 . 8 560
3.0 519
0.7 1 25
3. 1 140
4.8 505
2.8 480
2.0 200
2 . 0 200
1 .9 200
4 . 0 900
2.6 433
0.2 0
3.5 400
1.9 200
4.3 90
4.3 90
4 . 7 303
1 . 6 300
4 . 2 303
8.0 0
7.6 700
0.9 144
4 . 5 273
3 . 2 200
3.5 447
0.1 5
O.2 550
0.2 778
T H
P/1 F/V
0 0
0 0
60 0
100 0
0 0
63 0
100 0
0 0
50 0
60 0
100 0
100 0
100 0
100 0
100 0
0 100
100 0
50 50
0 0
0 0
40 0
100 0
50 0
0 0
0 0
100 25
50 0
100 0
67 11
100 0
0 0
50 0
F/E1 F/CO
100 0
100 0
40 20
0 0
1 00 1 00
38 25
0 0
100 0
50. 50
40 20
0 0
0 0
0 0
Q 0
0 0
100 0
0 0
50 0
100 50
100 50
60 60
0 0
50 50
100 100
100 100
0 0
50 50
0 0
33 33
0 0
100 0
50 0
PAGE
RATES «**««*
F/HC F/2
100 N/A
100 N/A
40 N/A
0 N/A
100 N/A
38 N/A
0 N/A
100 N/A
50 50
40 20
0 N/A
0 N/A
0 N/A
0 N/A
0 N/A
100 N/A
0 N/A
50 N/A
50 N/A
50 N/A
0 N/A
0 N/A
0 N/A
1 00 N/A
0 N/A
0 N/A
0 N/A
0 N/A
11 N/A
0 N/A
100 100
50 50
1
*»***
F/E
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
50
20
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
100
50
-------
KEY TO AIR - WEEKLY ERROR REPORT
COLUMN LABEL
Stn #
Mech #
Recs.Added
Rec.Errs.
Error Percentage by Field
Average Counts/CO and HC
Pass/Fail Rates
P/l
F/V
F/E1
F/CO
F/HC
F/2
F/E
DEFINITION
License number of emission, testing stations
License number of emissions mechanic
Total records completed by individual licensed
emissions mechanic that have been edited by the
Optiscan system and edited by the computer
Number of records that were improperly marked
or out of range for any edited field
(Next five columns) Make/License/Date/Cost/Cyl
Percentage errors of omissions, multiple markings,
out of range, not read by scanner, etc. as marked
by individual mechanic
Mean CO and HC readings as obtained from the total
of all first emissions tests
(shown as percentage)
Pass first test emission levels
Failed visual inspection (before first test)
Failed standards on first reading of CO and/or HC
Failed standard on first CO reading
Failed standard on first HC reading
Failed for any reason on retest emissions levels
Failed emissions only (will apply to 1982 autos)
217
-------
CO
TOTAL
tEPORT
IS OF FIRST FAILURES
C 0
L 0 R A D O D
E P A R T
M E N T O
F HEALTH
REPORT DATE
09/28/81
AIR PROGRAM
MODEL
YEAR
& e s s x
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
TOTAL
TESTS
X X X X S
598
941
1142
1350
1806
2139
2450
1894
2530
2603*
3258
3515
2851
1602
.11
28720
TOTAL
RE-
TESTS
ff S S X 8
112
167
164
227
351
376
406
311
424
568
663
725
259
72
0
4825
FAIL
RATE
& 8 S 3
18.7
17.7
14.4
16.8
19.4
17.6
16.6
16.4
16. 8
21 .8
20. 3
20.6
9. 1
4.5
0.0
16. 8
1 1 FIRST
1 1 CO
1 1 MEAN
6. 1446
6.6371
6. 1226
7.0291
6.7433
6.8811
7.0251
6.1113
6.6302
5 . 274 1
5.3003
4.4539
4 . 5954
4.0347
M
5.8309
RETEST
CO
MEAN
3.2634
3.4808
3.2244
3.2233
3.0353
2.8915
3. 1054
2.5566
2.8460
1 . 9342
1 . 8332
1 .5774
1 . 1 529
1 . 0528
M
2.3740
CORED
1ST CO
46.890
47.555
47.336
54. 143
54 . 988
57.979
55.796
58. 166
57.073
63.327
65.414
64 . 585
74.912
73.907
M
59.286
MEAN
CO
REDN
2.9
3.2
2.9
3.8
3.7
4.0
3.9
3.6
3.8
3.3
3.5
2.9
3.4
3.0
M
3.5
II FIRST
1 1 HC
1 1 MEAN
1427.9
1296.5
1259.9
1 169. 1
1 105.8
1046.7
1014.5
937.39
880.06
754 . 74
672.22
588.69
498.64
526.21
M
865.27
RETEST
HC
MEAN
818.59
688.10
723 . 62
649.14
537.42
510.36
531 .84
466.55
446.49
367.98
332.76
287. 18
215. 17
1 74 . 76
M
437.23
HCRED
1ST HC
42.672
46.927
42.565
44 . 476
51.402
51 .241
47.574
80.229
49.266
51.244
50.498
51.217
56.849
66.788
M
49.469
MEAN
HC
REDN
609
60B
536
620
568
636
483
471
434
387
339
302
283
351
11
428
-------
PO
•-•
vo
PASS/FAIL RATES
% Failing First Test
FIGURE 1
90
28
26
24
22
28
18
16
14
12
10
8
6
4
2
n
k
•
-
»
»
>
»
•
!
I
I
|
!
I
I
I
I
pn
I
i
i
1
1968 1969 1970 1971
1972 1973 1974
MODEL YEAR-TOTAL
1975 1976 1977 1978
72,686 VEHICLES
1979 1980 1981
11/12/81
-------
1ST CO
READING
\///\
FINAL CO
READING
ro
ro
o
MEAN CO REDUCTION
7-1-81 through 11-12-81
FIGURE 2
PERCENT CO
1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979
MODEL YEAR-TOTAL 72,688 VEHICLES
1981
11/12/81
-------
IV)
ro
100
95
90
85
60
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
I CARS PASSING AFTER ADJUSTMENTS
11,209 Vehicles
PASS RATE
1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981
MODEL YEAR
11/18/81
-------
APPENDIX H
223
-------
The following text is from SPSS, Statistical Package for the
Social Sciences, Second Edition, by Norman H. Nie, C. Hadlai Hull,
Jean G. Jenkins, Karin Steinbrenner, and Dale H. Bent (McGraw-Hill,
copyright 1975, by SPSS, Inc. eighth printing), pp. 11-19. It is
included in the report to provide a more comprehensive description of
the capabilities of SPSS for individuals not acquainted with the
package. It is reproduced with permission of SPSS, Inc., 444 North
Michigan Avenue, Suite 3300, Chicago, IL 60611.
225
-------
AN INTRODUCTION TO COMPUTING WITH SPSS 11
reduction that locates fewer underlying dimensions (higher-order variables) out of a larger pool
of variables in which no distinction has been made between independent and dependent vari-
ables. Canonical correlation is in some respects a combination of the two alternate multivariate
techniques. It contains data reduction capabilities similar to factor analysis, but, having required
the user to divide the variables into two sets, also assesses the relationship between the two sets
of factors (called canonical variates).
In this way, the researcher is able to conveniently simplify and analyze the relationship
between a large number of independent variables and a large number of dependent variables.
More precisely, canonical correlation analysis takes as its basic input two sets of variables, each
of which can be given theoretical meaning as a set, and extracts linear combinations of the
variables within each set; each linear combination maximally correlates with a corresponding
linear combination from the other set. These linear combinations are the canonical variates and
come in associated pairs. Thus, the higher-order dimensions are created not on the basis of
accounting for the maximal variance within one set of variables (as in factor analysis), but on
the basis of accounting for a maximum amount of the relationship between the two sets of
variables.
Input to the SPSS CANCORR procedure can be either raw data or a correlation matrix.
The user may specify the number of pairs of canonical variates to be extracted and the signifi-
cance level required for extraction. The procedure automatically outputs the canonical correla-
tions, along with tests of their statistical significance, and the coefficients of the canonical
variates. CANCORR will optionally punch or write the values of the canonical variates for all
cases in the file. These variates can be reentered into SPSS as new variables on a subsequent
run.
We have described the principal statistical procedures available within the SPSS system.
It is important to realize, however, that these procedures can be executed in any sequence, or
repetitively in the course of a single run or session with the computer. Thus the user may elect to
perform some crosstabulations, do a multiple regression, and then do some correlations upon
the same file of data in a single run. Also, the procedures described share the general
capabilities of SPSS for file handling, variable manipulation, and so forth, so that they consti-
tute a sequence of steps available to the user in any order that makes sense in the context of the
problem. In Sec. 1.3 we discuss some of the general capabilities of SPSS that arc available in
conjunction with any statistical procedure the user may specify.
1.3 AN OVERVIEW OF THE OPERATION OF SPSS
In this section we present a summary of the salient capabilities of SPSS, together with
examples. In subsequent chapters these features, and the manner in which the SPSS system
executes them, are discussed in greater detail. For the moment our purpose is to give the user an
overview of how the system operates, and to inform him of what he can and cannot accomplish
with it.
1.3.1 SEQUENCING CALCULATIONS
SPSS is driven through its various functions by a sequence of control cards1 that the user
must prepare. The process is illustrated in Fig. I.I. There is a control program in SPSS whose
sole function is to read control cards, decode them, and cause the appropriate function called for
by the control card to be executed. The control program causes the function to be performed by
passing control to the appropriate subprogram, which then performs the function and passes
control back to the control program, which then reads another control card, etc. This calculation
sequence is carried out automatically by SPSS, and the details of how the control program and
subprograms operate need be of no concern to the user. The important thing for the user to
'Throughout this text, the word card is taken 10 refer to an SO-churucter record recogni/ahle hy the computer. In
addition to implying the usual meaning (80-column IBM-curd formal), curd may refer to card-image records entered via
a remote terminal, etc.
227
-------
12
STATISTICAL PACKAGE FOR THE SOCIAL SCIENCES
St. ill ,
__»
1 — 1 1
Rr.
ml
i..
il .1
II. il
III
—*.
Ib llllS 1
1'INlSll
C.Kll'
Y,.,
Ni i
fWf
irm
SPSS
lunction
Stop
FIGURE 1.1 Program sequencing in SPSS.
realize is that SPSS processes control cards in sequential order. It is up to the user to arrange the
control cards so that the system will perform actions in the intended order.
The control cards themselves must be prepared in a particular format so that they are
recognizable to the system. There are over 75 different types of control cards in all, and the rules
for preparing these cards arc discussed in detail in subsequent chapters. An attempt has been
made to define the format of the control cards so that they correspond closely to the way the user
conceives the problem at hand. Thus, the information entered on these cards consists of a
quasinatural language for the description of data-analysis procedures. In order to use SPSS, il
is necessary that the user learn this language. This is not as formidable a task as it may
sound, since an attempt has been made to define the control cards so that they all have similar
formats and a minimum of rules is imposed on the user. The user is free to choose names and
labels that arc natural to the problem at hand.
1.3.2 ENTERING AND PROCESSING DATA
Data may be entered into SPSS in a variety of ways. The simplest and perhaps the most
common way is to punch the data on cards and to enter these cards along with the SPSS control
cards, which instruct the system on the processing of the data. Some of the SPSS control cards
define and describe the data while other types cause specific calculations to be executed. Data is
organized within the SPSS system in units called Jlles. A file consists of the user's data along
with associated information (entered on SPSS control cards) describing and defining the data.
Once entered, any such file may be permanently stored for future processing as an SPSS system
,/?/(' on tape, disk, or other input-output medium.
In Example I.I, we show the data and the control cards that the user would have to
prepare in order to begin to analyze the data from a hypothetical study of the political party
preference of 20 college professors. In this example, the data has been punched on 20 cards (one
corresponding to the data record.of each professor) and is placed in the card deck directly
following the READ INPUT DATA card. The data has been prepared in fixed-column format so
that each item of information for each professor is entered in precisely the same position on his
data record. In this example, the faculty member's identification number occupies the first four
card columns of every case. Party preference occupies column 6, and so forth.
The data-definition control cards that provide the system with information describing the
data required for processing are enclosed within a bracket and so designated. The first of these
cards, the FILE NAME card, simply names the set of data for future reference. The user may
also provide an extended label for the data on this card. On the VARIABLE LIST card the user
names each of the variables in the file of data. These user-selected variable names become
permanently associated with the corresponding variables in the file; and all future processing is
accomplished by reference to these names. The type of the variables and their location on the
data records is specified on the INPUT FORM AT card1 and the number of cases (professors in
"Readers familiar with Fortran will recognize ihai the format specifications of the INPUT FORMAT card are a subset of
the Fortran formal list.
228
-------
AN INTRODUCTION TO COMPUTING WITH SPSS
13
0«»R>INOT 6IVEN/
SEX I*M«|MALE »«PMFENALE/
RELIGION UIPROTESTANT I2ICATHOLIC OIJEMISH (MOTHER
MISS INC VALUES PARTYPRF !•«••/ AGE 101
PARTVPRF SEX Ul
TABLES* SEX BY PARTVPRF
ItS
l.l
. PR INT FORMATS
r CROSSTAB*
? < OPTIONS
-v V STATISTICS
RIAO INPUT DATA
4
. ^
f 1912 C 41 N
IBS* N 26 N
2TB6 L IS F
2576 R SO N
1631 N 61 F
2159 L 11 F
2614 L 45 F
15B2 C 56 N
2222 S 17 F
176B N 45 M
2651 S 10 N
2B42 S 44 N
1B99 N 0 F
2011 C 18 N
2159 L 0 F
1975 L IS M
24BB C 42 N 2
2111 N 16 F 4
1111 L 29 N 4
^ 2296 L 19 F 1
SAVE FILE
FINISH
EXAMPLE 1.1 Control cards and data used to enter data, perform a crosstabulation, and save a file
of 20 cases and 5 variables.
this instance) is indicated on the N OF CASES card. The fact lhal the data arc to he entered on
cards is indicated on the INPUT MEDIUM control card. If the data were being entered into the
system from some other input medium, the INPUT MEDIUM card would have specified a
keyword other than CARD (TAPE or DISK, for example). If this were (he case, (he 20 data
cards would not have appeared in the deck as shown in Example I.I. The PRINT FORMATS
specifies the printing format of the variables and is required only when there are variables in the
file that contain nonnumeric characters.
The next three types of cards provide SPSS with additional information frequently used
during processing. These cards are optional, however, and need only be prepared if the user
wishes to take advantage of certain features available within SPSS. The MISSING VALUHS
card enables the user to designate up to three values for each variable in the file to be treated as
missing. These values are specially treated during analysis, and each statistical program has a
number of user-selected options for processing missing values. Given (he frequency of missing
data in social science research, this card is almost always prepared, although it is optional. The
optional VAR LABELS cards permit the user to associate an extended label with any or all the
variables in the file. These labels are automatically printed on all tables and reports where
applicable. The VALUE LABELS cards serve an identical function for the individual values of the
variables and are also optional.
The data-definition cards need be prepared and entered only once, and the information on
them can be permanently saved along with the data as an SPSS system file. The SAVE FILE
control card directly following the last data card causes one of these specially formatted system
files to be created on an output medium of the user's choice. Once a system file has been
retained, the information initially entered on these cards is automatically passed from the file 10
229
-------
14 STATISTICAL PACKAGE FOR THE SOCIAL SCIENCES
the system along with the data whenever processing is desired. System files may be created for
storage during any processing run by inserting a SAVE FILE card in the control-card deck.
Thus a special run to generate the file is not required. Furthermore, while SPSS system files are
permanent, they are not immutable, and updated files may be created on any subsequent run.
While there are many advantages (to be discussed later) to generating system files, the
user may continue to input the data directly from cards or from any type of raw-input-data file,
and submit the control cards required to define the data on each processing run, that being the
method used with most statistical programs. The control cards required to process the data from
cards, tape, or disk files are identical (in kind and number) to those required to create a file for
storage as an SPSS system file. In the latter case, the user simply inserts a SAVE FILE control
card before the FINISH card.
While one need define a file of data only once, a new set of calculations or tasks will be
defined on each processing run. The SPSS system is instructed in the execution of the statistical
computations by means of a set of task-definition cards. The task defined in Example 1 . 1 calls
for a single table to be computed (using subprogram CROSSTABS), crosstabulating sex by
party preference. The CROSSTABS procedure card activates the crosstabulation subprogram.
and the OPTIONS and STATISTICS cards provide the CROSSTABS subprograms with addi-
tional detailed specifications for building the tables. The OPTIONS card enables the user to
control the direction of the percentaging of the table, the processing of missing values, the
printing of labels, etc. The desired table statistics, in this case chi-square and the contingency
coefficient, are selected by number on the STATISTICS curd. Figure 1.2 reproduces the printed
output computed for this run.
The remaining four SPSS control cards in Example 1 . 1 serve simple but special functions
in the system. The RUN NAME card, which may contain any message of the user's choice.
identifies the run, and (he message contained on it is reprinted on the top of each page of printed
output generated by the run (see Fig. 1.2). The READ INPUT DATA card informs the system
that the user has finished defining the file and the first statistical task, and the system is ready to
begin reading the data into the. computer. The SAVE FILE card previously mentioned causes
the file to be permanently saved as an SPSS system file, and the FINISH card simply informs
SPSS that the current run or session is completed.
The example presented shows how to enter raw data into the SPSS system and retain it for
future processing while performing a calculation. On any subsequent run the user can retrieve
this file automatically, and the variable names, formats, labels, etc., originally entered by the
data-definition cards are passed from the file, along with the data, whenever processing is
desired. Beyond the obvious advantages of automatic storing and retrieving of what may be
CeGSil»»lA«li, 1NU IHtN S»Vt >N SPSS STSItN FIL6 02/16/7*
FILE HC STUDY ICtEtTION UAIE • OIM6/T4I SUKVtf Of F4CULTT P»HI»
• ••••••••••*.••. CeOSSTtftULtTION U F ••••••••••••
SI I BT P4RIYPHF POLITICAL P»«TT PREFERENCE
• •••••••••••••.•••••••••••••••••••••••••••••••• f
PMKPMF
COUNT 1
CJL PCI ISOCUL C NEW OEMO LIB6S11 CONSt»* aO«
I*EDII CRAT IlVt TOUl
SE>
10
52.6
1
*7.*
COLUMN 3 57 » 19
IOI1L 15.8 26.1 36.a 21.1 100.0
CHI SOU1BE • 5.762*1 4 I [H ) DEGREES OF fottUOX SIKNIFICANCf • 0.1227
CONTINGENCY COEFFICIENT • O.kdJO*
NU»Bfc» UF MISSING UBSEUVMIONS • 1
1
1
1
1
1
s
2
(.6.7
1
3).}
N
2
60.0
L
2
28.6
5
71.*
C
*
100.0
0
0.0
FIGURE 1.2 Crosstabulations table produced by example run 1.1.
230
-------
AN INTRODUCTION TO COMPUTING WITH SPSS 15
large amounts of complicated file-defining information, the processing speed from these system
files is faster than that achieved with raw-input-data files in BCD form.' The conversion of data
from BCD to internal or SPSS system file representation also greatly facilitates the user's ability
to permanently recode variables and generate new scales and indexes through variable transfor-
mations, without concern for card and column locations of the new or rccodcd variables.
In the following respect, the files are permanent, bul not immutable. The data, or any of
the documenting information, may be added to. deleted, or altered at the user's will, and a new
or updated file may be retained. Additional variables can be added to the file us well as
additional cases; labels may be added or altered; new variables or scales can be created from
existing ones; and documenting messages may be saved in the file. In short, the system file
becomes a permanent self-documenting entity, and the user need only remember the name of the
file and the order of the variables within. Even this information, if forgotten, can be retrieved
easily.
The most important aspect of system files is the potential effect they can have (if properly
used) on the interaction between researcher and data during day-to-day analyses. With a com-
plicated data file, it will take considerable time to prepare and debug the initial run that defines
the file. However, once this has been accomplished, massive runs taking a long time to plan and
prepare need not and probably should not be made. With a system file the researcher can begin
to explore particular themes and hypotheses, submitting frequent runs requiring little card
preparation, and thus the likelihood of control-card errors is minimi/.cd.
Now that we have saved the example file called FACSTUDY a run exploring the
relationship between religion and party preference, controlling for the effects of sex, can be
made with the control cards of Example 1.2. The GET FILE card (the only control card not
previously introduced) causes all the data and required information from the file named on the
card to be read into the computer.
1 18
• UN NtMt FIRST fXANPU PROCESSING FRUN A SYSHM Mil
GET FILE FACSTUDV
CROSSTAB* TABLES • RELIGION 8V KARTVPMF 8V S6«
OPTIONS J,S
STATISTICS 1,2,1
FINISH
EXAMPLE 1.2 Control cards required to produce crosstabulations
from SPSS system file saved on the previous run.
As shown in Fig. 1.3. ihe output from this run is two completely labeled subtables
displaying the relationship between religion and party preference for male and female faculty
members. Comparable runs using many different types of statistical procedures could he made
as the analysis progresses.
Each of the control cards discussed above, as well as many other cards which perform a
variety of different functions in the system, are presented in great detail in (he following
chapters, and the reader should not be at all concerned if he feels that he only partially
understands the procedures and functions already discussed. The sole purpose of this introduc-
tory section is to provide the user with a brief overview of the capabilities of SPSS. File
handling and other general capabilities of SPSS are briefly described in the following sections.
'For those unfamiliar with the terminology, BCD slunds for binary-coded decimal ami corresponds to the recording
scheme normally used to punch data onio cards. The BCD recording scheme is one in which a single curd column is
used io represent a single digit or character. The binary recording: scheme used in SPSS system files is also acceptable as
raw input to SPSS, bul is rarely encountered. Other recording schemes such as column binary or multiple punching
cannot be directly input to SPSS.
231
-------
16
STATISTICAL PACKAGE FOR THE SOCIAL SCIENCES
ri»lt «*H»U MIOCISSI** FROM * irjtlK Ml-
rui i »cimo» ICM4HON O»TI • 02/14/Mi
or r*cuirv »««»» >«tfi«tnc£j
•U.IGION
CONMOU.IIIC rO«..
ill
U I * I I 0 N 0^ • ••••••••••»
IV MKtTfRP rOUIICAl rA«TY MtErEMNCE
MM.C
rAS€ i or i
COUNf I
C04, HI HOT 111 C Hi* 0(110 UMKAl
M«f
mute*
CATMOUC
OfHM ,
uoo
2.«*
4.00
COLUMI
TOTAL
k
94.0
I
a 0.0
o
0.0
i
jo. a
a
o.o
s».o
TIM
e
i
Tl.O
0
0.0
I
20.0
4
44.0
CNI IOU« II •
CMMII><> * •
i.ooooo VIM
o.ioaoo
20.0
4 MMtt! Of nkMOOM
IOTAL
4
4O.O
I
20.0
20.0
to
100. 0
ri«»f u**n.i MOCISSIM« r«ox 4
FtCSIUOT IClftTION 04 J I •
nil
au U/T*
IU»»I» Of FACULTY WITT
MilBION
CCNtMUlM
HI. . •
•T r««tYF«f
MLUI • r
FMTV MEFERENCI
IU1EU»>
. court i
COi Kt IIOCIU C MM DIM IIMIU
IKUlf CHAT
N
OIM
2.0*
4.M
COtUM
TOTU
I
1M.O
0
•.0
0
0.0
I
II.t
2
44.T
0
0.0
I
It.)
i
*«.o
t
40.9
0
a.o
TOTM.
1
I
M.I
1
U.I
1
11.1
CHI sou«it • s. 04000 mm
CMIICII*; v • o.»2*is
coHfiMiHcr cotrricitNf • o.
NUMCK dr HliilHC 0(SE*V*TIONS
>1.4 100.0
4 DlttUS W riHDQM
O.JIM
FIGURE 1.3 CroHtabulatlon tablet produced by example run 1.2.
1.3.3 SUBFILES
Data entered into the SPSS system may be substructured into groups called subfiles.
Subfiles may be sampling points such as cities; they may be national samples in crossnational
survey research; they may consist of data from different time trials or experimental treatments.
Subfiles, then, have all of the characteristics usually associated with like samples in statistical
analysis. In the SPSS system, the subfile identifier variable (SUBFILE) can be used as a control
variable for those statistical subprograms calculating like-sample test statistics, such as analysis
of variance. The subfile structure may also be used for more general types of comparative
analysis whenever the researcher has two or more like samples. The same relationship, for
example, can be examined simultaneously in each of the subfiles.
Once the subfile structure has been created, individual subfiles may be selected for
processing, combinations of subfiles may be processed together, or the subfile structure may be
ignored altogether—in which case, the data is treated as a unified file. The user controls the
manner in which the subfiles are processed in each individual task. At the end of a run, the user
may request that cases in the file be sorted and that a new subfile structure be defined on the
basis of the outcome of the sort. In this way, the basic subfile structure of the file may be altered
at the user's will.
232
-------
AN INTRODUCTION TO COMPUTING WITH SPSS 17
1.3.4 MISSING DATA
It is a common occurrence in social science research to find that for one reason or another
it has been impossible to obtain a complete set of data for every case in the file. Such a situation
would occur if a respondent refused or neglected to answer a question on a questionnaire, or if
the response was not entered correctly on the data sheets. SPSS has a number of features for
processing such missing data. Each variable may have up to three values that are designated as
missing. The choice of these values is totally a matter of the user's discretion, and is used to
designate the reason why proper data has not been obtained. For example, the user may elect to
use the code 0 for not applicable. 8 for don't know, and 9 for refused to answer. These
missing-data indicators may be defined by the use of a MISSING VALUES control card and
retained with the other information in a SPSS system file. Each of the statistical subprograms
contains a number of options for processing missing data, and the user may select whichever
option seems best suited to the particular analysis situation.
1.3.5 RECODING DATA
In order to organize data for analysis, the user first determines the variables to be dealt
with. The term variable means a certain attribute which can be determined or measured, and it
must be carefully distinguished from the term variable value (or value), which means the
value determined or measured for a variable in a particular case. After listing the variables, the
user next decides the way values of each variable will be coded. When the data is to be
processed with the computer, the way the coding system is devised can make a substantial
difference in the ease with which the user can cause the computer to carry out the desired
computations.
Frequently, the coding system originally used to record the data is not the most convenient
for use in all parts of the analysis. A provision has been made in SPSS for the user to change the
coding system after entering the data in its original form into the system. The value of any or all
the variables can be changed at the user's will by means of the RECODE process. Selected
values of variables may be replaced with new values, and continuous variables may be classified
into discrete categories. The RECODE process can be used to temporarily alter values of the
variables in conjunction with a run of a particular statistical subprogram, or it may be used to effect
a permanent receding of variables in the file.
1.3.6 VARIABLE TRANSFORMATIONS
A wide variety of variable transformations can be accomplished in SPSS by means of
simplified Fortranlike statements constructed by the user. The allowable types of transformation
are two types: conditional or unconditional. The unconditional transformations, defined by
COMPUTE control cards, cause a new variable to be constructed from the values of other
variables. For example, the control card
i ta
COMPUTE t«6»C
causes a new variable, named A, to be defined. The values of this new variable are determined
by adding the values of the existing variables B and C. Conditional transformations are defined
by the IF card. The IF card enables the user to test if a certain condition is true; if it is true, a
transformation is performed. Thus, the control card
1 16
If ID tQ II t-8»C
causes SPSS to examine the values of the variable D. If for a particular case the variable D
assumes the value I. a new value for variable A is computed by adding the values of variables B
and C.
233
-------
18 STATISTICAL PACKAGE FOR THE SOCIAL SCIENCES
Transformations can be used to normalize or in some other way alter the distribution of
variables as well as to construct scales or indices from two or more existing variables in a file.
The transformation process, like the recoding process, can be used to create a permanent file of
transformed variables, or it may be used to create temporary variable modifications during a
given run of SPSS.
A large number of repetitive transformations can be executed with a minimum of card
preparation by using the special repeat transformation. Sums of the occurrence of certain values
across a number of variables may be efficiently computed by means of the COUNT transforma-
tion card. Finally, a standard procedure for handling and assigning missing values for variables
created by any of the above types of variable transformations may be accomplished by the
ASSIGN MISSING command.
1.3.7 SAMPLING, SELECTING, AND WEIGHTING DATA
A random sample of the cases in a file may be obtained, specific cases may be selected
for processing, and the cases in the file may be weighted. The user is able to specify all the
conditions and criteria for accomplishing sampling, selecting, and weighting during any process-
ing run. As with recoding and variable transformations, sampling, selecting, and weighting
may be done in conjunction with a particular computation, or a new file of sampled, selected, or
weighted cases may be obtained.
1.3.8 AGGREGATING DATA
Research may involve dual levels of analysis or at least the examination of the impact of
some larger unit or institution on the behavior of individuals. Subprogram AGGREGATE
permits the researcher to define larger agfiregaiinn units and to compute aggregated vari-
ables.
Cases may be sorted into aggregation units on the basis of the values of any variable(s) in
(he file, Aggregated variables summarize the characteristics of the individuals in each aggrega-
tion unit; these variables arc the means, standard deviations, percentages, etc., of variables in
the file which have been measured at the individual (i.e.. lower) level. The aggregated variables,
along with aggregation unit identification numbers, are then punched on cards or written on tape
or disk, thus forming a new aggregated file.
Such aggregated files can then be input to SPSS or to other statistical programs, thus
shifting the level of analysis. Alternatively, subprogram AGGREGATE can produce a set of
aggregate characteristics which can be subsequently joined to the individual's data records in the
original tile. In this way, the user may perform what has come to be known as contextual or
compositional analysis.
t\
1.3.9 FILE MODIFICATION AND MANAGEMENT
SPSS makes available to the user a large number of general housekeeping routines for the
management, manipulation, and modification of data files. During any processing run. new
variables not currently in the file may be added to the file, providing a powerful facility for
merging separate sets of information on the same cases. Similarly, variables deemed to be of
no further use to the researcher may be deleted. Additional cases may be added to the file as they
become available to a study. Whole new subfiles may be added in a similar way. The cases of a
file may be resorted at the user's command for the purpose of aggregation or for the definition of
a new subfile structure. Master or archival files containing unusually large numbers of variables
can be accessed, merged, and saved, as well as converted into normal SPSS system files for
statistical processing.
Specified cases may be printed at the user's request. Missing-data indicators and a variety
of labelling information pertaining to individual variables or sets of variables may be altered or
234
-------
AN INTRODUCTION TO COMPUTING WITH SPSS 19
updated at the user's request. Finally, for the IBM-user community a special feature has been
added, enabling SPSS to read and convert data files distributed in the University of Michigan
Survey Research Center's special OSIRIS format.1
1.3.10 RETRIEVAL OF DATA FROM THE SYSTEM
All data input into the SPSS system, as well as recoded variables, new variables created
by transformation, and file changes accomplished by sampling, weighting, and/or selection may
be punched on cards or written in BCD form oh a device of the user's choice via the WRITE
CASES procedure. The cards or the BCD data files can then be directly input into the user's own
programs or into other statistical packages. The user has complete control over the selection of
variables to be output and their formats, as well as control over the selection of cases to be
output.
File-definition information, such as variable lists, labels, and missing-value codes may be
punched on cards in SPSS control-card format by the WRITE FILEINFO procedure, thus
facilitating the movement of SPSS files from one type of computer to another.
The SPSS correlation programs permit the user to output correlation matrices in BCD
form on cards, tape, or direct-access devices. All the SPSS multivariate routines using correla-
tion coefficients allow the user to input correlation matrices as well as raw data. In addition to
saving the user machine time by bypassing (he initial correlation step of these multivariate
techniques, the matrix-input feature allows matrices generated by the user's own program or
by other statistical packages to be input into SPSS subprograms; matrix output allows for
convenient use of matrices by non-SPSS programs as well as by those in the package.
Finally, Z scores, regression residuals, factor-scale scores, and canonical variables may be
output. Such variables may be subsequently merged with existing files.
These features enable the user to utilize the file management, data modification, and
statistical capabilities of SPSS without becoming a prisoner of the system.
1.3.11 OUTPUT OF RESULTS FROM THE SYSTEM
Since in this general discussion of SPSS and its features we have been concerned mainly
with problems of entering information into the computer, the user may well wonder how output
is obtained from the system as well. For the most part, printout is provided automatically so the
user need not be concerned with how SPSS accomplishes this. Generally, output occurs when
the user calls for a particular statistical procedure to be performed upon a file of data. The
system then causes the calculations to be made and produces a printed report containing these
results on the line printer or other output device. The level of detail that these reports contain
depends upon the level of detail provided when the file was defined.
For example, the user may define a variable named POLPREF. The user has the option of
also defining an extended label for POLPREF, as well as extended labels for the various values
which POLPREF can assume. If these labels are present, they will appear automatically on the
output reports. If they are not present, no labelling information appears. The user decides in
each case whether it is worth the trouble to enter additional information to make the printed
output more fully documented. -
There are some subprograms, such as the crosstabulation subprogram CROSSTABS,
which allow the user a good deal of latitude in specifying the level of detail to be contained in
the printout. The user may elect to use an OPTIONS control card to cause a subprogram to
produce the report desired. These options are discussed in those sections specifying the rules for
using various subprograms.
'OSIRIS is another widely used package of computer programs designed lor I he analysis ul social science data. The
system was developed by the Institute for Social Research, University of Michigan, and the Inter-University Consortium
lor Political Research.
235
-------
REFERENCES CONSULTED
Arthur Young & Company. Colorado Air Pollution Control Division
Information System (draft for review and correction by sponsor).
n. p.:Arthur Young & Company, 7 August 1981. Section 2.5.8.
Colorado Department of Health, Air Pollution Control Division, Mobile
Sources Section. Data Management for the AIR Program, July 1981
to December 1981. Denver, Colorado:Colorado Department of
Health, 20 January 1982.
Colorado Department of Health, Air Pollution Control Division, Mobile
Sources Section. Supplement to S.B. 52. Implementation Schedule
and Uork Plan for the "Auto Inspection Readjustment Program."
Denver, Colorado: Colorado Department of Health, 23 May 1980.
Colorado Department of Health, Air Pollution Control Division, Mobile
Sources Section. Various in-house documents. Various dates and
undated.
Fawcett, Rick. Unpublished inter-office communication to Jerry
Gallagher (Colorado Department of Health, Data Services Section).
5 August 1980.
National Center for Vehicle Emissions Control and Safety. Study Guide.
Colorado AIR Program. Fort Collins, Colorado: Colorado State
University. January, 1982.
National Computer Systems, Inc. Forms Design Guide. Reference Manual.
Sentry Optical Mark Reading Systems.Edina, Minnesota:National
Computer Systems, Inc.1980.
National Computer Systems, Inc. "Trans-Optic Custom Forms Price List,
Effective January 1, 1982." Minneapolis, Minnesota: National
Computer Systems, Inc. 1981.
Nie, Norman H., C. Hadlai Hull, Jean G. Jenkins, Karin Steinbrenner,
Dale H. Bent. SPSS. Statistical Package for the Social Sciences.
Second edition, 8th printing.Mew York:McGraw-Hill Book Company,
1975.
Radian Corporation. Guidance on Data Handling and Analyses in an
Inspection/Maintenance Program. Final Report.Austin, Texas:
Radian Corporation.1 December 1981.
Remnenga, Elmer. "Sample Size and Sample Plan Considerations for
Evaluating Emission Test Records for Consumer Protection."
Unpublished report to Colorado Department of Health. 9 October 1980.
Remmenga, Elmer. Unpublished memorandum to Jerry Gallagher, Industrial
Sciences [sic] (Colorado Department of Health, Mobile Sources
Section). 27 August 1980.
237
-------
|