United States
Environmental Protection
Agency
Mobile Source
Enforcement Division
Technical Support Branch
Washington DC 20460
November, 1978
&EPA
Motor Vehicle
Tampering Survey (1978)
-------
021835
1978
MOTOR VEHICLE
TAMPERING SURVEY
U.S. Environmental Protection Agency
Mobile Source Enforcement Division
Technical Support Branch
November, 1978
-------
First Printing November, 1978
Second Printing December, 1978
-------
TABLE OF CONTENTS
EXECUTIVE SUMMARY
I. Introduction
II. Design of the Survey
III. Tampering Inspection and Tampering Rationale
IV. Site Data
V. Major Results
VI. Further Results and Analysis
VII. Tampering and Inspection-Maintenance Programs
VIII. Emission Effects of Tampering
IX. Future Plans
X. Conclusion
APPENDIXES
-------
Executive Summary
The Technical Support Branch of EPA's Mobile Source
Enforcement Division has completed a survey which 1) assesses
the rate of vehicle emission tampering on a nationwide basis,
2) assesses the types of tampering most prevalent, and 3) quantifies
the relationship between tampering and idle test emission
scores. During the period May to August, 1978; 1,953 vehicles
comprising model years 1973 through 1978 were inspected at six
sites. These sites, Delaware, Maine, Virginia, Washington,
Tennessee, and Texas, were chosen to represent various national
geographic categories. Vehicles were observed while participating
in a non-voluntary program such as a mandatory state safety
inspection. The sites were selected such that the sample
reasonably represents vehicles not currently subject to an
emission oriented inspection-maintenance program.
Employees of the Technical Support Branch working with an
expert consultant and NEIC personnel conducted the inspections.
The same consultant was used at eacn site to assure a consistent
application of the inspection criteria.
At each site 200 to 400 vehicles were inspected for tampering
of 15 emission related components. In general, each inspection was
no longer than five Minutes. In addition to component inspection,
-------
idle hydrocarbon and carbon monoxide emissions were measured and
recorded. Fuel samples were taken from vehicles which required
unleaded gasoline.
The results of the survey classified each car into one
of four mutually exclusive groups: tampered (at least one
case of obvious tampering), arguably tampered (potential, but not
clear cut tampering), malfunctioning (some component not properly
functioning), or "O.K." (all components properly functioning).
The results show an overall tampering rate of 19% of the 1973
through 1978 vehicles in the nation. An additional 48% showed at least
one item in the arguably tampered category (e.g., limiter cap removal)^
JTh,
These results are shown in Table 1, which displays the degree
of tampered, arguably tampered, malfunctioning, and "O.K." cars
by model year.
Table 1 - Tampering Summary
Age Model Tampered Arguably Malfunctioning O.K.
0-1
1-2
2-3
3-4
4-5
5-6
Year
78
77
76
75
74
73
7%
10%
18%
22%
33%
32%
Tampered
31%
44%
53%
53%
57%
56%
4%
1%
1%
2%
1%
3%
58%
45%
28%
22%
9%
9%
Total 19% 48% 2% 31%
-11-
-------
There is a direct relationship between the tampering rate
and the age of the vehicle as shown in Figure 1. The projected rate
of tampering increases to approximately 50% as the fleet approaches
100,000 miles. The rates of tampering at different sites were
all within 5% of the 19% national average.
I The rate of tampering correlates well with increased idle
emissions scores. For each model year under consideration, idle
emissions were significantly higher for tampered cars than for
"O.K." cars./ EGR tampering and vacuum spark retard tampering
were the most prevalent forms of tampering. High idle carbon monoxide
levels were found to be indicative of NO related EGR tampering.
X
This is a secondary effect, however, indicating that a person
who tampers with the NO control system also tampers with the
A
idle settings. A relationship was also found between limiter
cap removal and higher idle carbon monoxide readings. Significantly,
72.5% of cars with limiter caps in place were below a 1% idle
CO level, while among vehicles with missing caps only 39.1% were below 1%
CO. The positive relationship between idle scores and tampering
demonstrates the ability of an idle test to detect tampered vehicles,
and a corresponding benefit of an inspection-maintenance program.
-111-
-------
i
!-••
<
T
i caca.ia
9ta.ca
BB.CB
•70.13
Bta.ca
SEB.BI
HB.BI
2H.CJ
ita.ta
Figure 1
Plot of Tampering vs. Age
(N=1953)
Y - 3.627 + 5.574(X)
R2= .9550
F = 84.95
5 s a ? a
fin
a a
-------
From the subset of vehicles inspected which required
unleaded gas, an overall fuel switching rate of approximately 5%
can be estimated. This is the percentage of vehicles with leaded
fuel in the tanks (as determined by sample analysis) and/or tampered
filler neck restrictors. Somewhat surprisingly, 75% of the cars
with evidence of leaded gas in their tanks did not have tampered
fillernecks, indicating either alternate methods of filling a car
with leaded gas (e.g. funnels, slow fill), or nozzle violations at the
gas pump. Conversely, 62% of the vehicles with tampered filler neck
restrictors did not have leaded gas. Presumably, this indicated only
sporadic use of leaded gas, or some other rationale such as second
ownership of the vehicle. In either case, whatever motivated
the removal of the filler neck did not apparently motivate
continued switching. The 6% fuel switching rate differs from the
estimate of 10% found in an MSED gas station observation survey.
Differences in the vehicle mix observed in each survey may explain
some of the discrepency . In order to assess the reasons for these
differences further sampling will be instituted concentrating
on 1977 and 1978 model year vehicles. It takes a minimum of three
unleaded fills on top of a new empty tank (1 gallon left) to lower
the tank lead level to less than .05g/gal lead after one fill of
leaded @ 2.0g/gal. of lead .
The emission effect of an overall tampering rate of 19% is
significant. Approximate calculations indicate that by the
-v-
-------
time a vehicle reaches 50,000 miles, the average hydrocarbon and
carbon monoxide emissions will be nearly four times that of a
vehicle with no tampering or malmaintenance.
-vi-
-------
I. Introduction
The f'.ooile Source Enforcement Division of tne L'.'J. El'A
is responsible for enforcement of the mobile source provisions
of the Clean Air Act, as amended August, 1977. Sections
203(a)(3)(A) and 203(a)(3)(B) of the Act prohibit the removal
or rendering inoperative of emission controls by dealers and
the service industry, and are generally referred to as the
"taripering" provisions of the Act.
Prior to 1378, the Division had collected evidence
indicating that tampering was occurring. However, these
data were very difficult to quantitatively assess Because of
the variability anong inspection procedures and inspectors.
In conjunction with this variaoility was the ausence of a
specific method by which a non-voluntary vehicle sample
could oe inspected. The qualitative data that was available
originated nostly from I/A areas and a voluntary sample, and
was assuned to reflect less tampering than woulcl be expected
in non-I/M areas and with a non-voluntary sanple. A sunr.iary
of these data is presented in Table 1.1.
In early 1973, the Technical Support Branch undertook a
systematic national tampering survey of non-I/Li areas in
-1-
-------
Table
III. Mobile Source In-Use
Year Site/Study
1977-78 New Jersey1
1977-78 Fuel Switching
(Nationwide)
1977-78 VW Field Fix
Survey
1977-78 Ford Recall Data
1977 Portland, Oregon
1976-77 CARB
1976-77 MSAPC Restor.
Maint.
1976 Chrysler Recall
Data
1975 MSAPC Emission
Factors Program
1975 New Jersey
1974 New Jersey
1974 D. C.
1973 EPA Rpt. Auto.
Exh. Emission
Surv. (1973)
1 Preliminary results
2_ Rabbits and Sciroccos, 19
3_ 3 of 15 randomly selected
4_ Checked only for catalyst
restriction
c \ w
•H £ -0
11 a £ AJ
J"± US U M ^'^
W Qt -H O -H dt/l
o e e e oj e H
it m -H to 3 o e
Vehicle MY1 s ° * J * r* u *
1975-78
1975-1978
1975-762
1977
1975-77
1975-1976
1975-76
1975
1972-76
1970-1976
1970-1974
1970-74
1968-71
14%
—
100%
20%3
8%4
15%
-30%*
24%5
—
6%
11%
15%
—
51%
—
__
—
—
—
45%
— —
—
57%
53%
33%
—
—
10%
__
—
—
—
—
—
—
—
—
—
—
—
—
__
—
—
47%
42%
—
23%6
—
—
—
37%7
of 19 randomly selected vehicles
vehicles
, unleaded labels, fuel inlet
5_ 7 of 29 randomly selected vehicles
6 By city: Denver-7% Los Angeles-33% Che. Houston,
St. Louis, Phoenix, D.C.-31%
By model year: 1976-47%, 1975-28%, 1974-13%, 1973-13%, 1972-17%
7 HC and CO only; By city:
Denver 6%, By model year:
1969-70%, 1970-81%, 1971-62%.
* Unconfirmed data
-2-
-------
order to obtain data sufficient to (1) quantitatively project
the incidence of tampering on a national level with high
confidence, (2) to determine the most prevelant types of
tampering, (3) and to determine the relationship between
tampering and idle emissions.
The original strategy was to "tag-on" to mandatory
state/city safety inspection programs where a representative,
non-voluntary vehicle sample would be expected. This strategy
was subsequently modified to include "tagging along" on
state police license/safety roadside checks for the same
reason. The strategy called for one expert to supervise the
inspection teams throughout the survey, in order to minimize
variability.
II. Design of the Survey
The major elements of design were as follows:
1. A minimum of six (6) sites in areas with mandatory
safety inspections would be required for pro-
jection to a national level.
-3-
-------
2. 300 to 500 vehicles per site would yield the
1800 to 3000 vehicles necessary for a confident
national projection. Survey dates were to be
chosen for peak flow such as at the end of the
month when inspection stickers expire.
3. 1973 and newer light duty vehicles and light duty
trucks would be the target population. 1973 was
the first model year that these vehicle classes
possessed significant emission controls.
4. The sample would be completely random. No attempt
would be made to approximate the national vehicle
mix according to manufacturer, model, or model
year. The main goal was to survey as many vehicles
as possible, and not skip any because they were
rare or too many of one type had already been
inspected.
5. Representative geographic diversity would be
essential and sites would be chosen from major
geographic areas of the country.
6. The inspection must be thorough and check all
visible emission control systems, record basic
-4-
-------
data, obtain a fuel sample, and require less than
four to five minutes. It was felt that this speed
would be necessary if the cooperation of the
safety agencies was to be maintained. A detailed
description of the inspection is given later
in this report.
7. At least two inspectors would be required, and
in the busiest cases, three or four.
8. Idle HC and CO emissions would be measured with
portable analyzers.
9. A single expert should perform the inspection
of the major emission control devices throughout
the survey in order to maintain consistency. This
person must have a thorough knowledge of the
emission control systems used by the major domestic
and foreign manufacturers.
10. The inspection would objectively record the con-
dition of the emission controls. A method would
be developed for determining which conditions
constituted tampering.
-5-
-------
11. The survey should be completed by September, 1978
An "expert" in emission controls was recruited from the
private sector to conduct the summary. Jack Gockel, a
recognized emission expert, was under contract to MSED and
agreed to perform the survey.
In arranging survey sites, the agencies responsible for
the vehicle safety inspections in the following areas were
contacted:
New York State
Georgia
Delaware
Nebraska
Chattanooga, Tenn.
Memphis, Tenn.
Washington, D.C.
Virginia
Florida
Texas
Not all non-I/M areas with mandatory safety inspections
were contacted. The above were judged to be the most desirable
from which to select a geographically diverse sample. Contact
-6-
-------
was initially made by phone. Letters requesting permission
to conduct the survey were sent to each along with a list of
the items to be inspected in each car. It was emphasized
that this was an informational survey and that no enforcement
action would be taken.
It should be noted that the term "tampering" was avoided
in oral and written communication with the agencies. The
survey was described as a "check of the emission control
equipment."
All agreed to participate except Memphis and Florida.
Memphis has a high volume 5 minute safety check where a
vehicle stops for one minute at each of 4 positions along
the safety lane. Memphis was concerned that the tampering
inspection would be too disruptive to the flow. Florida
declined because of current controversy in the state
legislature over the state safety inspection. The Florida
state police did not want the safety inspection to become
involved in any activity that could result in adverse
publicity.
Delaware, Tennessee (Chattanooga), Texas, and Virginia
were chosen as survey sites because they represented a
-7-
-------
reasonable geographic diversity. However, it was judged
that at least two additional sites in other areas of the
country were necessary. Thus, additional states were contacted
concerning the possibility of "tagging along" on state
police roadside license and safety checks. Using this
approach, roadside surveys in Maine and Seattle, Washington
were arranged.
In principle, participation in the tampering survey was
voluntary. That is, the driver was given a choice before
his car was inspected. However, by virtue of the safety
inspection or roadside check by police, the audience was
"captive" or non-voluntary and thus most people were encouraged
and inclined to participate.
III. The Tampering Inspection and Tampering Rationale
The circumstances under which the sample was obtained
in each area will be described in a later section. The
following is a description of the inspection procedure, and
the rationale used to determine tampering. The forms used
to record data (figures 3.1 and 3.2) were essentially identical
to forms used by the New Jersey Department of Environmental
-8-
-------
FIGURE 3.1
HPA Vehicle Tampering Study
Date:
-------
-10-
FIGURE 3.2
EPA Vehicle Tampering Study
Codes:
Jisconnectcd Unit
~Tl octrical - 1
Vacuum - 2
Mechanical - 3
loses
Incorrect Rt - 4
Rev.Operation -5
Bypass - 6
For Is
Missing - 7
Hi sadjus ted - 8
Malfunction - 9
Functioning
Properly - 0
lion-Stock Equip.-
99
-------
Protection in performance of a prior tampering survey for
MSED. These forms covered all the items to be checked in
the tampering inspection, and did not need modification
except for the addition of space to record odometer readings
The functional check of emission components was performed
mostly by Jack Gockel.
Arrangements were made with the Region III lab in
Wheeling/ West Virginia to analyze the fuel samples for
lead content.
A. Equipment
The following equipment was used:
1. Leaded nozzle to check the fuel filler inlet for
penetration.
2. Hand fuel pump with a 3 ounce sample bottle
and approximately three feet of hose.
3. Flashlight.
4. Hand held mirror.
-11-
-------
5. Hand vacuum pump and gauge for use in the EGR
functional check, including hose and adapting
nipples.
6. Fender covers.
7. Screwdriver and pliers for disconnecting and
reconnecting vacuum hoses.
8. Two Horiba Mexa 300 A emission analyzers
9. Power converters to run the analyzers off the
car batteries when out in the field.
10. Gasoline can for replacement fuel and fuel to
flush the fuel line after each sample.
11. Spare fuel line.
12. Extension cords.
13. 200-300 sample bottles per site. (1973 and 1974
vehicles did not require fuel samples.)
14. Boxes and tape for packing fuel samples.
-12-
-------
15. Labels for fuel sample bottles.
16. Shipping labels to Wheeling, W.Va. fuel analysis lab,
17. Flammable liquid labels for fuel shipment,
18. Clip boards.
19. Spare data forms,
20. Spare parts for the fuel pump.
21. Spare fuel pump,
22. Spare leaded nozzle,
23. Pens for recording data (black ink, not blue)
24. Cloth gloves.
E. Data for Vehicle Identification
The following data were recorded on the form presented
in figure 3.1.
-13-
-------
1. Date
2. Vehicle identifying survey number - vehicles were
numbered sequentially as they were inspected,
and this number was followed by a site identifying
letter. For example, "S" was used for the Seattle
area.
3. Make - determined by observation.
4. Model - determined by observation.
5. Model year - obtained from driver and verified by
underhood emission label.
6. Odometer mileage
7. Engine family/CID as recorded from the underhood
vehicle emission control information label.
8. HC in ppm and CO in % for idle and high rev
(approximately 2500 rpm).
9. The presence of smoke.
10. Carburetor - if the carburetor was original
equipment a "P" was used to indicate that it was
a production unit. If fuel injection was used then
"FI" was recorded. If the carburetor had been
replaced with a non-stock unit, then a "99" was
used. Jack Gockel's experience was relied upon
to determine if a carburetor appeared to be a
non-stock unit. The number of barrels was easy
to determine by the size and shape of the carburetor
-14-
-------
body. A one barrel carburetor had a one "jug"
body and one idle mixture screw. Two and four
barrel carburetors had two idle mixture screws
and two and four "jugs" respectively. The
number of barrels was recorded under "carb
model".
C. Emission Control Checks, Data Recording, and Tampering
The form presented in Figure 3.2 was used to record the
inspection of the emission control devices. The following
codes, as shown on the form, were used.
Code
0 - item is functioning properly
1 - electrical disconnect
2 - vacuum disconnect
3 - mechanical disconnect
4 - incorrectly routed hose
5 - hoses connected so that an item operates in a reverse
manner (special case of incorrectly routed hoses,
mostly applicable to vacuum delay valves).
6 - hoses routed such that an item is bypassed (another
special case of incorrectly routed hoses).
7 - missing item
-15-
-------
8 - misadjusted item
9 - malfunctioning item
99 - non-stock equipment
The codes are designed so that the inspector can ob-
jectively record the condition of a device, and he does
not have to make an "on the spot" judgement with respect to
tampering.
The observations were broken down into five categories:
tampered, arguably (or questionably) tampered, malmaintenance,
malfunctioning, and "O.K." Most of the items considered
tampering are obvious. However, some are not so obvious and
the rationale for placing them in the tampering category is
presented. Figure 3.3 shows how the codes were categorized
for each item. Note that not every code applies to each
item.
No malmaintenance, defined by figure 3.3 as a misadjusted
air injection system or vacuum spark retard, was found. Thus,
the results discussed in Section V, Major Results, will
consist of only four categories, "O.K.", tampered, arguably
tampered, and malfunctioning.
The tampering rationale is based on the following
sections of the Clean Air Act: Section 203(a). The following
-16-
-------
-17-
FIGURE -3.3
EPA Vehicle Tampering Study
Codes:
D\sconncctod Unit
Tlf.'ctrica! - 1
Vacuum - 2
Mechanical - 3
Hoses
ncorrect Rt - 4
Rev.Operateon -5
Bypass - 6
Farts
Missing - 7
Mi sadjus ted - 3
Malfunction - 9
^ Q Functioning
gigf Properly - 0
3' S //OA' ^
Equip.
99
SPECIAL
CASE
AIR CLEANER
TAMPERING
-------
acts and the causing thereof are prohibited —
(3)(A) for any person to remove or render inop-
erative any device or element of design installed
on or in a motor vehicle or motor vehicle engine
in compliance with regulations under this title
prior to its sale and delivery to the ultimate
purchaser, or for any manufacturer or dealer
knowingly to remove or render inoperative any
such device or element of design after such^sale
and delivery to the ultimate purchaser; or
(B) for any person engaged in the business of
repairing, servicing, selling, leasing, or trading
motor vehicles or motor vehicle engines, or who
operates a fleet of motor vehicles, knowingly to
remove or render inoperative any device or element
of design installed on or in a motor vehicle or
motor vehicle engine in compliance with regulations
under this title following its sale and delivery to
the ultimate purchaser.
The "knowingly" element does not imply intent to tamper,
but rather knowledge that an act occurred. For example,
some people have inverted the top of the carburetor air
cleaner under the assumption that it will improve engine
-18-
-------
performance. This renders inoperative the positive crank-
case ventilation system, the evaporative control system in
most cases, and the heated air intake system, and while it
may not be done to tamper per se, it is none the less a
prohibited act.
The following items were inspected. (See figures 3.2
and 3.3). The applicable codes and inspection criteria are
noted, and tampering is discussed.
1. Idle stop solenoid - This solenoid provides an idle
stop for maintaining idle speeds to the higher speeds
needed to minimize CO emissions. On some vehicles,
it is used to close the throttle and thus prevent
run-on when the engine ignition is turned off. On
vehicles with air conditioning, it is used for in-
creasing engine idle speed to compensate for a
decrease in idle speed when the air conditioner is
engaged.
(a) The electrical connection was checked.
If it was disconnected or broken - code 1.
A disconnected connection is considered
obvious tampering. A broken connection,
-19-
-------
could have been deliberate or accidental.
Both cases were considered tampering because
it is unlikely they could have occurred except
by a deliberate act.
(b) With the air conditioner on, (or in non-air
conditioned vehicles) the solenoid should
activate and contact the throttle linkage.
With the air conditioning turned off there
would be a small gap between the solenoid
stop and the throttle linkage. If the
appropriate condition was not observed, the
solenoid was recorded as malfunctioning -
code 9. Time did not permit a check to see
if the solenoid was defective or just mis-
adjusted.
(c) Only codes 0, 1, 8, and 9 were observed.
(Only one code 8 was observed, and it was
grouped with code 9.)
2. Heated air intake - provides warm air to carburetor
during cold engine operation.
-20-
-------
(a) If the vacuum line to the vacuum override
motor (top of air cleaner horn) was dis-
connected or Hissing, then code 2. It
normally requires a deliberate act by a
mechanic or owner to disconnect this line,
and thus, it is considered tampering even
though failure to reconnect might be due
to negligence.
(b) If the "stovepipe" (the flimsy olack paper/
foil connection between the exhaust manifold
shroud and the air cleaner horn) was
missing - code 7. If it was present but
not properly connected, (hanging from one
connection, for exanple) it was considered
disconnected - code 3. If it was torn or
deteriorated, it was also considered dis-
connected - code 3. u>out half of the ve-
hicles had metal stovepipes that were not as
susceptible to these problems.
(c) If any problems v/ere evident with the butterfly
plate in the air cleaner horn which appeared
to restrict its movement, it was considered to
be malfunctioning - code 9.
-21-
-------
(d) If the air cleaner top had been inverted, or
holes were punched in the air cleaner, the
heated air intake was considered disconnected
- code 3. This is considered tampering because
the heated intake is deliberately defeated
and it also causes the PCV system and some
evaporative control systems to malfunction.
Malfunctions are registered through the code
9's on the PCV and on some storage canisters
which purged with a line connected to the
air cleaner.
(e) The question of tampering for a mechanically
disconnected or missing stovepipe (codes 3 and 7)
is controversial. Certainly an improperly
connected, missing, or torn stovepipe renders
the heated intake inoperative. However, no
one who understands engines would expect any
improvement in performance or fuel economy due
to its removal or disconnection. These problems
are the result of fragile design and careless
replacement of the air cleaner, not of any attempt
to tamper. Many times the foil just fits over
the sharp edge of the exhaust shroud, and if it
is not pushed on carefully, it easily tears. The
-22-
-------
incidence of codes 3 and 7 is so high that it was
determined that it would be misleading to place
them in a tampering category. Thus, codes 3 and
7 are labeled as arguably tampered. If code 3
is coupled with code 9 on the PCV, this indicates
the altered air cleaner, and is tampering.
(f) Problems with the butterfly plate in the air
cleaner horn are considered defects, and are
labeled as malfunctions.
(g) If a non-stock (code 99) air cleaner filter was
installed and this element was wider than the
production design, a gap was created in the air
cleaner chamber. In this case, a 9 was recorded
for the PCV system, a 9 for some storage canisters,
and a code 3, mechanical disconnect, for the
heated intake. This is considered tampering for
the same reason as the inverted air cleaner top.
3. Limiter caps - plastic caps on idle mixture screws
designed to limit carburetor adjustments.
(a) The inspector checked for the presence of the
caps. In most cases they were missing - code 7.
-23-
-------
(b) If caps appeared in good shape - code 0.
(c) If the caps appeared damaged, they were
checked to see if they could be rotated
past the stops. Caps with broken or bent
tabs (allowing rotation past the stops) were
given a 3. In some cases, the data recorder
designated a 9. In this case, the code 3 and
code 9 designate the same condition, a mech-
anically disconnected and ineffective
limiter cap.
(d) Limiter cap removal is prevalent. To place it
in the tampering category would obscure the
rest of the data. However, it has defeated
an element of emission control design and is
done knowingly. Thus, missing or disconnected
limiter caps are considered arguably tampered.
This misadjustment usually causes a significant
increase in CO emissions. Mechanics remove
limiter caps to enrichen the idle mixture for
smoother engine operation. Enrichments pro-
ducing greater than about 1% CO in the exhaust,
do not provide enough oxygen for the correct
oxidizing function of the catalyst. As a
-24-
-------
result, the vehicle usually exceeds EPA standards,
These misadjustments are actions which render
catalysts inoperative and represent the greatest
single cause of maintenance related air pollution
from post-1974 vehicles.
4. Positive crankcase ventilation system - A typical
configuration for a V-8 engine consists of the PCV valve
connected to a valve cover and then connected to the
carburetor by a vacuum line. The other part of the system
has a "fresh air tube" running from the air cleaner to the
other valve cover.
(a) If the line between the PCV valve and the
carburetor was disconnected then code 2 des-
ignates a vacuum disconnect. If the "fresh
air tube" was disconnected, code 3 was recorded.
Some code 4's, incorrect routing, were also
recorded for this condition.
(b) Missing valve and hoses - code 7.
(c) A code 9, malfunction, was recorded if the
air cleaner top was inverted, if holes had
been punched in the air cleaner, a too thick
-25-
-------
air filter was used, or if there were any
holes or tears in the PCV lines.
(d) For add-on devices a non-stock equipment
"99" was recorded.
(e) The PCV system is difficult to categorize
for tampering. A good mechanic would not
defeat it because it is common knowledge
that it improves engine life by purging the
crankcase of blow-by gases. In many cases,
it becomes disconnected when the air cleaner
is removed and it is inadvertently left that
way when the air cleaner is replaced. How-
ever, it is also the target for many "gasoline
savings" devices that install in the vacuum
hose leading to the carburetor base. The PCV
system is an important control that a) prevents
crankcase emissions, b) is clearly visible,
and c) is easily maintained. Thus, codes
2,3,4,7, and 99 are considered to be tampering.
Code 9 is considered tampering only when it
results from a tampered air cleaner.
-26-
-------
5. Evaporative control system - Controls vapors from
the fuel tank and carburetor. Some systems have two lines,
one from the fuel tank to the- canister, and one from the
canister to the carburetor or air cleaner to air purge the
canister. Other systems have a third line which is usually
connected to the carburetor.
(a) The lines were checked for proper routing and
connections. If any were missing or disconnected,
code 2 was recorded for vacuum lines and code 3
was recorded for others. Both cases are clearly
tampering.
(b) If the canister was missing, code 7 was recorded.
This is clearly tampering.
(c) As discussed under the PCV system, if there
were problems with the air cleaner, then
code 9 was recorded if one of the canister
lines was connected to the air cleaner.
6. Tank cap
(a) The cap was checked for a good gasket that
sealed properly. If it did not seal properly,
-27-
-------
code 9 was recorded. If the cap was missing,
code 7 was recorded. (One code 3 was found, and
it was grouped with code 9.)
(b) The gas cap is also a controversial tampering
item. Its absence defeats the evaporative control
system. Yet, it is highly unlikely that it would
be left off on purpose. It was decided that
there is no excuse for the cap to be missing, and
that it should be considered as arguably tampered.
A bad gasket is considered a malfunction.
7. Air injection system - Consists of an air pump driven
by a crankshaft pulley which pumps air through a control
valve and lines connected to the exhaust manifold.
(a) The pump, belt, lines, and control valve were
checked visually. The most common problem was
a missing drive belt which was usually recorded
as code 7 for "missing". Occasionally, the data
recorder improperly designated a code 3 for
mechanical disconnect. Both codes designate the
same condition. Removal of the drive belt
renders the pump and the control valve inoper-
-28-
-------
ative. Therefore, a code 9 for "malfunctioning"
pumps and control valves was recorded in most
cases.
The second most common air system problem was
the removal of the air pump. This was denoted
by a code 7 for both the pump and the valve. In
many cases, the control valve is an integral part
of the pump - thus, a 7 for the control valve.
In a few cases, the recorder used 3 for mechanical
disconnection for designating this same condition.
Other more minor problems were the disconnection
of hoses and lines. If the vacuum line was
disconnected to the control valve a code 2 was
used. If the inlet or outlet hoses to the pump
were disconnected a code 3 was used.
(b) Codes 2,3, and 7 were considered blatant
tampering. While there were inconsistencies
in recording the codes for the individual
components, an accurate percentage of total
system tampering may be determined if the
system is analyzed as a whole.
-29-
-------
8. Exhaust gas recirculation system - The standard con-
figuration consists of a vacuum line from the carburetor
to a sensor (used to detect temperature to activate the
EGR valve), and another vacuum line from the sensor to the
EGR valve. Some systems have multiple sensors and thus
additional vacuum lines. The system directs a portion
of the exhaust gases back into the cylinders for the con-
trol of oxides of nitrogen. This is one system where a
functional check was performed. EGR valves are found in
sealed and non-sealed units, and require different function-
al checks.
Non-sealed EGR valve functional check
(a) The system was visually inspected to see if
the valve, sensor(s) and hoses were in place.
If any hoses were disconnected, code 2 was
recorded for the associated valve or sensor.
Missing parts received code 7.
(b) The engine was revved and EGR valve stem
movement was checked visually or by touch.
If stem movement was detected, code 0 was
entered for both the valve and the sensor,
and the system passed the check.
-30-
-------
(c) If no stem movement was detected, the vacuum
line to the valve was pulled off and checked
for blockage. If it was blocked, code 2 was
recorded for the valve. If it was not blocked,
a hand vacuum pump was then connected to the
sensor outlet and the engine revved. If a
vacuum was obtained, the sensor was good and
the valve was given a 9 for a malfunction. If
no vacuum was obtained, then the line to the
sensor from the carburetor was checked for
vacuum while revving the engine. If this line
was blocked, then code 2 was recorded for the
sensor and code 9 for the valve, if the valve
did not already have a code. If this line
to the sensor had vacuum, then the sensor was
bad and code 9 was recorded for both the sensor
and the valve.
(d) Some lines had a vacuum delay valve in them
that was positioned backwards. A reverse
operation, code 5, was recorded.
(e) On some systems with multiple sensors and
multiple nipples, some code 4, mis-routings,
were found.
-31-
-------
(f) A crushed nipple on the EGR valve or sensor
resulted in a code 3 for that component.
(g) For systems with multiple sensors, each sensor
and hose was checked as above if no stem
movement was detected.
Sealed EGR valve functional check
(a) The system was visually inspected.
(b) The vacuum hose to the EGR valve was discon-
nected. The hand vacuum pump was connected
to the valve and vacuum applied with the
engine running. If idle speed dropped with
the application of vacuum, the valve was
good. The vacuum pump was then inserted
into the line leading to the valve's vacuum
source. The engine was revved to determine
if vacuum was available. If vacuum was
available, the sensors were checked using
the same procedure described for the non-
sealed unit.
(c) Obviously, all codes except 9 are tampering
(no code 6's were observed). Note, if a sensor
-32-
-------
was bad, for what ever reason, the EGR valve
was coded 9 automatically and not usually checked,
Thus, the number of malfunctioning EGR valves
is the total of the EGR valves that failed the
functional check and the EGR valves assumed to
have malfunctioned due to a bad or tampered
sensor. The main reason for not checking each
EGR valve after finding a bad sensor was time.
However, EGR valves tend to freeze with dis-use
and it would be impossible to determine which
were defective and which froze from dis-use.
9. Catalytic converter
(a) The converter was visually inspected for its
presence and high temperature discoloration.
The only applicable code was "missing" - code
7. This is obvious tampering. If there was
any uncertainty as to whether or not a vehicle
was supposed to have a catalyst, the underhood
vehicle emission control information label was
checked. On most vehicles, the emission control
systems are listed on this label. Chrysler is
a notable exception.
-33-
-------
(b) It is possible to remove the catalyst material
from all General Motors' catalysts through a plug
in the body of the canister. Our experience
indicates that some catalyst material removal can
be expected. However, an inspection for this
was not performed in this survey.
10. Dash labels and tank labels
(a) The only applicable code is 7. Removal of the
labels, which are an element of emission control
design, is considered arguably tampered.
11. Filler Neck - Unleaded vehicles only.
(a) The filler neck inlet restrictor was checked
with a leaded nozzle. If it fit (whether or
not alteration was visible) a code 3 was
recorded. If the restrictor was missing -
code 7.
(b) It was difficult to determine which cars
require unleaded fuel. Catalyst cars definitely
required unleaded fuel. However, now several
manufacturers produce non-catalyst cars that
-34-
-------
require unleaded fuel. The easiest method to
clarify any doubts was to check the dashboard
since that label is rarely missing.
(c) Both acts are considered tampering. (Code 3
and 7.)
(d) One unleaded pick-up truck had an auxiliary fuel
tank installed with a leaded inlet. This was
recorded as a non-stock 99, and was considered
tampering.
12. Vacuum Spark retard system.
(a) Visually checked for proper connections.
(b) Codes 2 or 3 - disconnected - tampering.
(c) Missing - 7 - tampering.
(d) Only codes 0, 2, 3, and 7 were observed.
13. Tampering Source - with time permitting, some
drivers with tampered cars were asked if they knew the
origin of the tampering. Little value is given to
these observations.
-35-
-------
0 - Owner
D - Dealer
OD - Other than dealer
DK - Don1t know
D. Inspection Scenario
A typical inspection required a four-man crew and proceeded
as follows:
The driver was approached and greeted with:
"Good morning, we are with the U.S. Environmental
Protection Agency and we are performing a survey to
check the condition of emission control equipment.
If you have a few minutes, we would like to look
under your hood and see if everything is in good
shape, and also take a 3 ounce fuel sample to check
for lead content."
The general answer was "yes" or "How long will it
take?"
Those who questioned the time were told "three or four
minutes." Most who were apprehensive at first were placated
by the short time required. If the driver asked directly if
-36-
-------
he was required to participate, he was informed that he
wasn't, but that it was an important survey and it would be
greatly appreciated if he would. Some people pleaded a case
of being late and left. It is estimated that less than 5%
skipped the tampering inspection. One reason for not
participating used several times was that the driver was not
the owner of the car and that he didn1t want anything done
to the car without the owner's permission.
Once the driver agreed, the car was brought to the
inspection area and the hood was raised. On unleaded
cars, one team member would check the dash unleaded
fuel label, the unleaded label at the filler inlet, the
size of the inlet with the leaded fuel nozzle, to see
if it would fit and push open the flap, and visually
inspect the inlet, tank cap, and tank cap gasket. He
would then draw out a fuel sample.
The sample bottle was then tightly capped, marked with
the survey identifying number, and placed in a container for
shipment to the analysis laboratory. When requested by the
owner, approximately 3 ounces of unleaded fuel was poured
into his tank to replace the sample. A small amount of new
fuel was then poured through the hand pump line to flush any
leaded fuel residue that might have remained.
-37-
-------
Some problems were encountered with locking gas caps.
If the driver had to shut off his engine to get the key to
open the cap and several vehicles were waiting, the locking
gas caps on leaded cars were not checked.
When a damaged or missing gas cap gasket, a tampered
filler inlet, or missing labels were found, it was recorded
on the forms.
As soon as a car arrived at the inspection point, one
team member would place the analyzer probe in the tailpipe
and begin to fill out the vehicle data form. When all the
vehicle data and inspection information except for the
emission values had been obtained, he would read and record
the analyzer at idle, then at a high rev. He would then
remove the probe.
Meanwhile, the consultant, using a flashlight and
mirror, would call out his underhood observations to another
team member who recorded them on the appropriate form. When
the observations were complete and Jack had revved the
engine to complete the emission measurements the car was
sent on its way. Time permitting, we would discuss any
tampering we had found with the driver. Most drivers were
very nice, cooperative, and interested. The few that were
very adamant were sent on their way without an inspection.
-38-
-------
IV. Site Data
A. Maine - Roadside Check.
Dates -
May 16 through May 19, 1978
Participants -
Maine State Police
Jack Gockel (Expert Consultant)
Mel Petroccia - EPA Region I
Randy Rice - EPA Region I
Robert Knowles - EPA Region I
Samples -
338 vehicles
Fuel samples -
Maine refused to allow fuel sampling
for reasons of safety.
Sites -
Tuesday - highway in Bangor suburbs.
Wednesday - under an overpass in
downtown Bangor.
Thursday - weigh station on highway
between Gay and Springfield.
Friday - same as Thursday.
-39-
-------
The State Police conducted roadside registration and
safety checks and passed the vehicles along to the survey
team. The drivers were made aware of the tampering survey
by the police and encouraged to participate. The police
were very helpful and interested in the tampering survey.
The tampering survey was rained out on Tuesday afternoon and
a sheltered site was arranged for Wednesday, and it rained
most of the day. Thursday was cloudy and Friday was sunny.
Temperatures were in the 60's° F.
B. Virginia I - Private garage safety inspection.
Dates -
May 30 through June 2, 1978.
June 19 through June 21, 1978
Participants -
Jack Gockel
Paul Gesalman - EPA/MSED
Jim Caldwell - EPA/MSED
Steve Albrink - EPA/MSED
Truman Wilson - NEIC
Samples -
98 vehicles (May 30-June 2)
13_ vehicles (June 19-June 21)
111 Total
Fuel Samples -
39 analyzed
-40-
-------
9 destroyed in shipment via UPS.
Sites - Fairfax Exxon, 10480 Lee Highway,
Fairfax, Va. (May 30-June 2).
Exxon at Seven Corners, Arlington,
Va. (June 19-21).
Weather - Hot, sunny, 80's° F.
Virginia has a semi-annual private garage safety in-
spection. Flow is very low compared to a roadside check.
However, the State Police initially refused to allow us to
tag-on to a roadside check and that left only the garage
sites. The peak inspection flow period is the final week of
the month since the inspection stickers run out the first of
the next month. Jack Gockel and Paul Gesalman worked the
May 30-June 2 period and averaged about 25 vehicles a day,
with the station inspecting around 40 to 50 vehicles per
day, including pre-73 models. The low flow rate in this
type of inspection requires only two inspectors. Because of
extremely low flow rates, the inspection was discontinued after
the third day of the second week and the State Police were
contacted regarding a roadside check. A roadside check was
arranged.
-41-
-------
C. Virginia II - roadside check.
Dates -
Monday, July 24-July 28, 1978.
Participants -
Virginia State Police
Jim Caldwell - EAP/MSED
Paul Gesalman - EPA/MSED
Larry Walz - NEIC
Ron Snyder - NEIC
Steve Sinkez - EPA/MSED
Samples -
313 vehicles
Fuel samples -
189 obtained and analyzed.
Paul Gesalman drove the samples to
Wheeling.
Sites -
Monday - Parking lot of a dance hall
on rural but busy State Route 17
south of Warrenton. Afternoon
sampling was rained out.
Tuesday - weigh station on U.S. Route 1
in Woodbridge - very busy road.
Wednesday - State route 234, five
miles south of Manassas.
-42-
-------
Thursday - Route 1, five miles south
of Alexandria, in high school
parking lot.
Friday - same as Thursday.
Weather - Hot, mostly sunny, 80°F.
One trooper was assigned to perform the check each day,
He was asked to send us only 1973 and later vehicles with
Virginia plates. A few out of state cars were inspected.
The trooper's inspection was very quick and there was gen-
erally one car waiting for the tampering inspection at all
times.
D. Wilmington, Delaware - State run safety lanes.
Dates - June 12-16, 1978.
Participants - Jack Gockel
John Fahrenback - EPA/MSED
John Davis - EPA/MSED
Janet Littlejohn - EPA/MSED
Charles Case - EPA/MSED
Paul Gesalman - EPA/MSED
Jim Caldwell - EPA/MSED
-43-
-------
Samples -
360 vehicles.
Fuel samples -
56 analyzed
83 are questionable because they
were returned by UPS in poor condition
and upon analysis, gave very atypical
results.
37 destroyed by UPS.
176 Total
Site -
Wilmington suburb, near airport.
Weather -
Hot, sunny, 80°F, moderate humidity,
Delaware has an annual safety inspection performed at
four state operated sites. The site chosen for an inspection
was the largest and had four lanes, only two of which were
being used. The team used a vacant lane outside the inspection
building and cars were chosen by the team randomly from
those waiting in the other lanes. The safety inspection
personnel were very accomodating. In most cases, the vehicles
we took out of the safety inspection waiting line were let
back in behind the car originally in front of them. The
flow was very heavy at times and overall just as good as the
roadside check.
-44-
-------
E. Chattanooga, Tennessee - City run safety lanes,
Dates -
June 26-30, 1978.
Participants -
Jack Gockel
Paul Gesalman - EPA/MSED
Tom Newman - NEIC
John Dion - NEIC
Samples -
325 vehicles.
Fuel samples -
23 analyzed
98 lost by UPS.
Weather -
Hot, sunny, humid, 90°F, high air
pollutant levels.
Chattanooga has a city-wide annual safety inspection
at one four-lane site. Approximately 500 vehicles are
checked per day in a four minute inspection. The tampering
inspection was performed in front of an active lane and
appropriate vehicles were directed to that lane by safety
inspection personnel. The tampering survey received local
TV and press coverage.
-45-
-------
p. Seattle, Washington - Roadside check.
Dates -
Monday, July 10-July 14, 1978
Participants -
Washington State Police
Jack Gockel
Bob Bernstein - EPA/MSED
Larry Walz - NEIC
Ron Snyder - NEIC
Jim Caldwell - EPA/MSED
Samples -
323 vehicles.
Fuel samples - 161 were taken and analyzed.
Sites -
Monday - Rural highway in Seattle
suburbs.
Tuesday - busy roadway in industrial
area.
Wednesday - intersection of two inter-
state highways near Seattle.
Thursday - exit ramp of an interstate
highway in Seattle suburbs.
Friday - local park in Seattle suburbs
Weather -
Hot, sunny, 80°F, low humidity,
-46-
-------
A roadside check was conducted by the Washington State
Police. The police performed a license/safety check and
sent the vehicles to the tampering inspection team. There
were generally two or three troopers present. The assistance
from the troopers was excellent. Region X was very interested
in the survey, sent several observers including a local
politician, and arranged local TV and press coverage.
G. Houston, Texas - Private garage safety inspection.
Dates -
Monday, August 28 - Saturday,
September 2, 1978
Participants -
Jack Gockel
Paul Gesalman - EPA/MSED
Herbert Vaughan - EPA/MSED
Ron Snyder - NEIC
Larry Walz - NEIC
Samples -
218 vehicles.
Fuel samples -
89 obtained and analyzed.
Sites - Team 1 -
Monday - Mobile Station in a depressed
neighborhood of downtown Houston.
There were many old vehicles.
-47-
-------
Tuesday - Exxon station on highway in
a Houston suburb.
Wednesday - Texaco station in affluent
suburbs.
Thursday - Service station in downtown
Houston.
Friday, Saturday - same as Thursday.
Team 2 -
Monday - same service station team 1
surveyed on Thursday.
Tuesday - Saturday - Shell station in
suburbs.
Weather -
Hot, sunny, humid, 90°F.
An attempt was made to conduct a roadside survey in
Texas. However, the Texas Department of Public Safety did
not feel it would be appropriate to delay traffic beyond the
roadside license/safety check. In order to increase the
sample size obtainable at private garages, two tampering
inspection teams were formed. The State police took the
teams to the stations and stayed until the tampering in-
spections got under way.
-48-
-------
Jack Gockel lead team 2. Paul Gesalman lead team 1 and
was replaced by Larry Walz for Thursday - Saturday.
-49-
-------
V. Major Results
This section presents the basic aggregate results of
the survey. Total tampering by site, model year, and type
of tampering are discussed. As the data for each vehicle
were processed, the vehicle was classified into one of four
categories: "O.K.", tampered, arguably tampered, and mal-
functioning. Since each vehicle inspected has various
components, each of which could be tampered, the vehicle itself
is classified by the worst state of any component in the vehicle.
The hierarchy is in the order: tampered, arguably tampered,
malfunctioning, not equipped, properly functioning. Thus, if any
one component is tampered, the entire vehicle is considered
tampered. If one component is "arguably tampered" and all the
others are functioning properly, the entire vehicle is considered
"arguably tampered". Thus an "O.K." vehicle must have all
components functioning properly.
An "O.K." vehicle is one with no observable gross tampering,
arguable tampering, malfunctioning, or missing equipment. The
term, "O.K.", does not mean that the vehicle is necessarily operating
properly. For example, a spark plug or coil may not be performing
satisfactorily resulting in a poorly operating and excessively
polluting vehicle. This situation would still be classified "O.K."
for purposes of this tampering report.
-50-
-------
£. Aggregate Results
The results for all vehicles tested are presented in Appendix
V-I and summarized in Table 5.1
Table 5.1 - Aggregate Results
Percent Number
O.K. cars 30.7% 599
Tampered Cars 18.9% 370
Arguably Tampered 48.4% 945
Malfunctioning cars 2.0% 39
100.0% 1,953
These results span several model years and the totals here
represent a snapshot of vehicles at a specific point in time
(Summer 1978). As such, the numbers must be viewed with caution.
For instance, it would be incorrect to conclude that a mere 2.0% of
the vehicles had malfunctioning components. This number appears
low because of the hierarchy explained above. Many cars had com-
ponents malfunctioning but if at least one other component were
tampered or arguably tampered, the vehicle, as a whole, would
have been classified in the tampered or arguably tampered state.
Thus as soon as an item like a missing limiter cap (arguably tampered)
is noted, the car cannot be classified as malfunctioning, no
matter how many other components malfunction.
A major concern is the degree of confidence that exists
in the 18.9% tampering estimate. Because of the large size of our
sample, we are 95% confident that the true proportion of tampered
cars is between 17.1% and 20.7%.
-51-
-------
B. Tampering by Vehicle Age
Appendix V-l also presents a breakout of tamperi.no categories
by model year of the vehicle. It is notable that the rate of
tampering increases with the age of the vehicle, and correspondingly,
the percentage of "O.K." cars decreases with the age of the
vehicle as shown in Table 5.2.
Age
—•^k«B
0-1
1-2
2-3
3-4
4-5
5-6
Tampered
7.4%
10.1%
17.7%
22.3%
32.6%
32.0%
O.K.
58.1%
44.6%
28.4%
22.3%
9.4%
9.1%
Figure 5.1 demonstrates the virtually linear nature of the increase
in tampering rate over the life of the vehicle.
This tampering increase with age is significant. The 19%
tampering rate represents an average tampering rate in calendar
year 1978. The data trend shows that by the end of six years
in the life of a model, 37% of the cars will be tampered with.
After eight years of driving (approximately 100,000 miles)
about 48% of the fleet will be in the tampered classification.
This means that almost half the cars will be in the tampered
group, not just malmaintained or arguably tampered.
-52-
-------
tn
i cscj.ta
sca.ca
BEJ.EJ
•7G3.GJ
Bca.ca
srca.ta
HG3.GJ
EH
20.13
i ta.ca
§ aca.ia
Figure 5.1
Plot of Tampering vs. Age
(N=1953)
Y = 3.627 + 5.574(X)
R2= .9550
F = 84.95
1 — 1 1
•
1 ID K| ™
a — " m ri
, 1 H- 1
?aa S| B!
lA in r-
i 1 •
n ni —
Age (in years)
-------
Total cars
306
416
216
324
335
356
Tampered
15.7*
14.2%
22.2%
20.4%
20.9%
22.8%
Tampered
41.8%
49.8%
59.7%
44.4%
46.3%
51.1%
Malfunctioning
2.0%
2.6%
1.43
.9%
2.11
2.5%
O.K.
40.5%
33.4%
15.7%
34.3%
30.7%
24.2%
C. Tanpering by Site
Appendix V-2 breaks out tampering by site. It can
be summarized in Table 5.3
Table 5.3 - Tampering by Site
Arguably
Site
Washington
Virginia
Texas
Tennessee
Maine
Delaware
Total 1953 18.9% 48.4% 2.0% 30.7%
It can be seen that there is some variation from the 18.9%
nationwide average in the different sites. In an effort to
ascertain if these departures from 18.9% are significant or if they
are the result of random fluctuations, a Chi-Square test at the 5%
level of significance was run. The results proved just significant,
meaning randon chance could have resulted in these fluctuations
only 5% of the time.
Interestingly, when the same test was run on the percentage
of "O.K." cars; the results proved very significantly different.
There are less than 5 chances in 1000 tnat such a wide discrepancy
occurs due to random chance. The factor that appears responsible
for this is the low number of "O.K." cars in Texas.
-54-
-------
D. Tampering by Type of Vehicle
In order to ascertain if any vehicle make exhibited more
tampering than others, a breakout by make was run and is shown
as Appendix V-3. This has been summarized by manufacturer in
Table 5.4:
Table 5.4 - Tampering by Manufacturer
Manufacturer Total cars Tampering
GM 882 20.1%
Ford 491 20.2%
Chrysler 242 20.1%
AMC 64 31.3%
Foreign 272 8.5%
A Chi-Square Test was also run on these numbers. The
results were significant at a 5% level. This indicates a
distinct difference in tampering level amongst manufacturers.
This difference is probably due to the low level of tampering
emong foreign vehicles.
E. Summary
To summarize, the national rate of tampering in 1978 is
18.9%. Tampering increases with age, and as many as 50% of the
vehicles which are 8 years old can be expected to have been
tampered with. The tampering rate appears higher for domestic
makes than for foreign makes, and tends to be affected by
geographic location.
-55-
-------
VI. Further Results and Analysis
The previous section dealt with the major aggregate
tampering results by age, site, and manufacturer. In this
section some further insight is provided by examining specific
individual forms of tampering, as well as certain "groups"
of tampering such as fuel related tampering, (fillerneck
restrictor tampering, gas cap tampering, and/or excess lead
levels) or EGR system tampering (EGR control valve or EGR
sensor tampering). Further, the relationship between idle
emission levels and tampered vehicles is discussed.
A. Individual Types of Tampering
The rates of tampering by type are given in Table 6.1
Categories narked with an asterisk (*) were not considered
as tampering.
Table 6.1 - Individual Rates of Tampering
Idle Stop Solenoid 0.74%
Heated Intake 0.84%
Limiter Cap Removal *
PCV Valve 3.27%
ECS Storage 2.63%
ECS Tank Cap *
Airpump 3.17%
Air Control Valve 2.90%
Air Pump Belt 5.74%
EGR Control Valve 11.87%
EGR Sensor 5.32%
Catalytic Converter 1.21%
Dashboard Label *
Fuel tank label *
-56-
-------
Filler Neck Removal 3.37%
Vacuum Spark Retard 10.50%
Air Cleaner 0.87%
From these figures, it is apparent that EGR control and
vacuum spark retard tampering are the two major sources of
tampering. It should also be noted that while limiter cap
removal was not included as definite tampering, this removal
occurred in 65% of the vehicles observed. Assuming that a
removed limiter cap allowed the idle to be set rich, the
effect on emissions of such an action could be huge.
Given the high rate of limiter cap removal, an
investigation was made on the effect of this on idle emission
scores. Appendixes VI-1 and VI-2 demonstrate the mean HC
and CO idle scores by model year for cars with and without
limiter caps. The summary is shown in Table 6.2.
Table 6.2 - Limiter Cap vs. Idle Scores
Model Year
73
74
75
76
77
78
Mean
O.K.
182
180
122
145
128
54
Idle HC (ppm)
Missing
297
297
188
259
188
101
Mean Idle
O.K.
2.36
2.21
.99
1.35
.81
.52
CO (%)
Missing
3.41
3.35
2.19
2.52
2.07
1.40
-57-
-------
It is quite clear that idle scores for both HC and CO
with limiter caps removed are considerably above those with
limiter caps in place for all model years. Further, of the
1844 cars that had limiter caps required as equipment, a
test was made to see if their idle CO would have exceeded
1%. The results were that 72.5% of the vehicles with
limiter caps in place would have passed a CO test with a 1%
CO idle cutpoint. Conversely, 60.9% of those who removed
the limiter cap would have failed such a test.
Since EGR tampering was the highest single item of
tampering, it was desired to observe the idle test's ability
to identify this form of tampering. Since EGR tampering
affects NO , one might expect little relationship between
A
idle HC and CO scores and EGR tampering. The results showed
differently however. As demonstrated by Appendixes VI-3 and
VI-4 average HC and CO idle scores increased as the EGR
system was tampered with. A summary is provided in Table
6.3
Table 6.3 EGR Tampering vs. idle scores
Model Year
73
74
75
76
77
78
Idle
O.K.
247
295
164
235
152
67
HC (ppm)
Tampered
348
321
206
236
205
51
Idle CO
O.K.
3.04
3.07
1.73
2.19
1.40
.71
(%)
Tampered
4.28
3.68
2.17
3.00
1.75
1.21
-58-
-------
Since, for physical reasons one would not expect an HC
or CO effect from NO tampering, it appears that vehicles
x
with NO system tampering also exhibit tampering with other
x
systems or with limiter caps and thus degrade their HC and
CO controls as well as their NO controls.
A
To assess the degree of this collateral effect, the
data were examined for the amount of NO tampering that
A
coincides with limiter cap removal, and the amount of N0x
tampering that occurs with at least one other forn of tampering
or arguable tampering.
The results showed that in at least 80% of the cases of
NO tampering, limiter cap removal also occurred; and in
A
86% of the 223 cases of NOX tampering, at least one other
form of tampering or arguable tampering was present. This
result tends to support the hypothesis that a vehicle exhibiting
NO tampering also exhibits other tampering. In fact, the
A
EPA Restorative maintenance study has demonstrated that even
if EGR disablement is the only form of tampering, resultant
HC and CO emissions will increase as well as the anticipated
increase in NOx.
B. Relationship of Tampering and Idle Emissions
Idle tests are the predominant test made associated
-59-
-------
with inspection and maintenance programs. To the extent
that idle test scores are capable of discriminating between
tampered and non-tampered cars, the more useful the scores
become in an inspection maintenance application. A few
examples have been brought to EPA's attention in which
extreme amounts of tampering including catalyst
removal, have not affected a vehicle's ability to pass an
idle test. These samples have shown the ability to circumvent
an idle test if other settings were modified accordingly.
This survey provided a unique opportunity to assess the
ability of the idle test to detect tampering on a large in-
use sample in non l/M states. Thus the results would not
be based on a handful of specially prepared cars, but rather
on a typical cross-section of in-use vehicles.
In order to assess this ability of the idle test, low
and high CO and HC idle scores were measured on each vehicle.
The following analyses have considered only the low idle in
assessing idle scores and tampering.
Mean idle scores have been calculated for each of the
four categories of cars (O.K., tampered, arguably tampered,
and malfunctioning) for each model year in the survey.
Appendixes VI-5 and VI-6 show the results. "O.K." cars have
-60-
-------
Generally lower idle scores on both HC and CO than either
the tampered or arguably tampered cars for every model year,
For malfunctioning equipment, the idle scores are comparable
for some post-1975 models.
The most important results are reproduced below in Table
6.4
Table 6.4 - Relationship of tampering and mean idle scores
Model Year Mean Idle HC(ppm) Mean Idle C0(%)
O.K. Tampered O.K. Tampered
73 124 328 1.95 3.95
74 230 313 2.63 3.43
75 115 211 1.07 2.49
76 151 254 1.43 2.66
77 109 221 0.687 1.77
78 56 51 0.483 0.591
In an analysis of variance test, idle scores proved to be
very highly significantly correlated to tampered vehicles.
C. Fuel Related Tampering
Previous information indicated that a proportion of the
population of vehicles requiring unleaded gas was being
fueled with leaded gas. To evaluate the extent of this
phenomenon, fuel samples were taken from vehicles requiring
unleaded gas. Due to some difficulties in transporting the
gas samples to laboratories, only 481 valid samples could be
-61-
-------
considered for the study. From this smaller sample an
attempt was made to relate filler neck inlet tampering to
the evidence of leaded fuel (in excess of 0.05 grams per
gallon) in the tank. The results appe'ar in Appendix VI-7.
Only 20 vehicles requiring unleaded gas (4.2%) had leaded
gas in the tank. Among these 20, only 5 had tampered
filler neck restrictors. This appears to indicate the use
of some alternate means of filling the vehicle with leaded
gas such as funnels or a slow fill procedure.
The percentage of the 481 vehicles with leaded fuel
samples and/or tampered filler neck retrictors was 5.8%.
The 95% confidence interval for this overall fuel switching
potential rate, based upon a sample of size 481 is from 3.6%
to 8.0%.
Since this estimate differs from another MSED study
showing approximately 10% fuel switching, further analysis
is being undertaken to examine any differences in vehicle
populations observed that might account for this discrepancy,
Further, more data will be obtained in order to attempt
reconciliation of the observed differences. The additional
sampling will concentrate on 1977 and later model year
vehicles.
-62-
-------
VII. Tampering and Inspection Maintenance Programs
In previous sections, the relationship between idle scores
and the extent of tampering has been discussed. The basic results
were that tampered cars had appreciably higher mean idle CO levels
and but for one model year had higher idle HC levels than
O.K. cars. Further, idle HC and CO scores were also found higher
for tampered NO related control devices such as the EGR system. The
A
probable reason for this is the concurrent tampering of the EGR
system with at least one other component affecting CO or HC, such
as limiter cap removal. Additionally, a very high correlation was
found between limiter cap removal and the inability to pass an
idle test with a cutpoint of 1% CO.
The next analysis was to simulate these vehicles undergoing
a New Jersey inspection - maintenance program. New Jersey
modified phase II cutpoints of SOOppm HC and 5.0% CO for model years
73 and 74; and SOOppm HC and 3.0% CO for model years 75 and later
were used. The results are summarized in table 7.1
Table 7.1 - Relationship between Tampering and New Jersey Program
Pass Fail
O.K. 86% 14%
Tampered 65% 35%
Arguably Tampered 68% 32%
Malfunctioning 80% 20%
-63 -
-------
As can be seen, 35% of the tampered cars failed the New Jersey
test, while only 14% of the "O.K." cars failed. Thus there appears
to be a higher probablity of failing such a test with a
tampered vehicle. It should be noted that the 65% of tampered
vehicles which pass New Jersey standards do not necessarily mean
that the idle test is incapable of identifying tampering, but
rather that the general New Jersey standards can be considered
somewhat lenient/ in that most cars, tampered or non-tampered,
appear to pass.
To summarize, of the O.K. vehicles 86% pass New Jersey
standards, while vehicles in one of the other three groups have
only a 67% chance of passing. The difference between the 86% and
67%, based on the sample size of this survey, is statistically very
significant, demonstrating a marked inverse relationship between
the rate of tampering and the ability to pass the New Jersey program.
It is important to remember that these relationships are derived
from studying a set of vehicles not currently subject to an emissions
inspection maintenance program. Any secondary effect such
as the mere presence of such an I/M program dissuading tampering
actions would not be reflected in these analyses.
-64-
-------
VIII. Emission Effects of Tampering
The 19% overall tampering rate, along with projections
for increased tampering by age for each vehicle model year
have a considerable effect on resultant emissions. To try
to quantify these effects, an emissions model which considers
each component with its tampered condition effect on emissions,
and the time phasing of different types of tampering is
required.
While such a complete model does not now exist, an
approximation may be made using an emission model as described
in "Emissions from Catalyst Cars Beyond 50,000 Miles and the
Implications for the Federal Motor Vehicle Control Program",
(SAE paper 780027). This model considers vehicles to be
classified in one of three states: properly functioning,
poorly maintained, or disabled catalyst. For purposes of
comparison, we have considered limiter cap removal as an
indicator of poor maintenance. Further, not all the tampering
observed in this study is of the extreme nature as catalyst
disablement. Thus, the limiter cap removal rate, which
increased by 10% of the vehicles per year was used as the
rate of malmaintenance. The tampering rate which increases
at 4% per year in a model's life was reduced to 3% to account
£or tampering incidents less egregious than catalyst dis-
ablement.
-65-
-------
If these estimates are used in the approximation model,
the results show a four fold increase in the average carbon
monoxide and hydrocarbon emissions in the fourth year of the
model's life. This means that by the time a fleet reaches
50,000 miles, the average emissions from this fleet will be four
times the average emissions of the same fleet having no malmainte-
nance and no tampering. This estimate is a fairly gross
approximation and is, of course, dependent on the assumptions
made in the model. However, it does provide a good "ball park"
assessment of the effects of tampering on vehicle emissions and
should serve to underscore the severity of the problem.
The effect of EGR tampering on emissions has not been studied
extensively, but some data is available. In a Restorative Maintenance
Program conducted by EPA thirty-seven 1975 and 1976 automobiles were
tested with their EGR systems deliberately plugged. Emissions increased
from the non-plugged condition by an average of 21%, 71%, and 123% for
HC, CO, and NO respectively.
X
-66-
-------
IX Future Plans
The rates of tampering discussed in this paper were
found in the period May - August/ 1978. The results represent
a rather large sample and reflect the condition of vehicles in
this time period. While specific future plans for tampering
assessment are still being designed, it is felt that the current
survey has adequately assessed the tampering situation at the
present time.
Thus, the thrust of future plans will be to observe the change
in tampering rates observed as programs and resources are implemented
to curb this problem. The specific recommendations are:
1. A survey in New Jersey of the same type
performed in the other six sites, utilizing the
same personnel, equipment, and definitions. In this
way a consistent comparison can be made between tampering
rates and the implementation of an inspection maintenance
program.
2. Future controlled surveys, probably using
contractor assistance, of sites in two EPA regions
per calendar quarter. This will serve as a monitor
of the trend of future tampering. For regions earmarked
for concentrated tampering enforcement, or which have
-67-
-------
on-going inspection maintenance programs/ this will
enable a measurement of the effects of these programs.
To provide for a consistent check of tampering
levels, it is suggested that the EPA consultant
currently supervising the surveys be utilized to train
MSED, contractor, and regional employees in the methods
of tampering and the categorization of vehicles.
3. Additional investigation in order to understand
the differences between the fuel switching rate
observed in this study and other studies.
-68-
-------
X Conclusions
A tampering study of 1,953 vehicles estimates the tampering
rate at 19% gross tampering in areas not currently subject to emission
inspection maintenance programs. Since the vast majority of the
nation is comprised of non - I/M areas*, this rate is considered to
be a reliable nationwide estimate. The tampering rate increases
with the age of the vehicle from about 5% for new vehicles to close to
50% for 100,000 mile old vehicles. The most prevalent forms of gross
tampering were the EGR system (13%) and the vacuum spark retard (11%).
In addition, very high rates (65%) of limiter cap removal were observed
Vehicles with limiter caps removed had idle CO in excess of 1%
most of the time (61%). The tampering observed has a very pronounced
emission effect. Fuel switching as assessed by this survey is
occuring in about 6% of the vehicles requiring unleaded gas.
Idle test scores detect tampered cars. In fact, idle test
scores even catch EGR tampering which is N0x related, due
to the high level of concurrent presence of EGR tampering and at least
one other form of tampering. This tends to strengthen the
usefulness of inspection maintenance programs due to their
ability to discourage vehicle tampering.
Current mandatory I/M areas are: New Jersey, Cincinnati,
Portland (Oregon), Arizona (two counties), Nevada, and
Rhode Island. Approximately 6% of the nation's vehicles
are included in such a program.
-69-
-------
The surveillance should continue with emphasis on comparing
these results with those in areas having an on-going inspection
maintenance program, and in investigating future trends in
tampering rates. The reasons for the difference between fuel switching
observed in this study and other studies should be established.
-70-
-------
Appendix V-I
-------
TESTING OF DATA-AND PROGRAM
FILE NONAME (CREATION DATE 6 09/25/78)
09/25/78
PACE 2S»
MYR
COUNT 1
R04 PCT ]
COL PCT
TOT PCT
73.
74".
J
J *
75.
76.
77.
78.
COLUMN
TOTAL
STATUS
OK
0.
23
9.1
3.8
1.2
26
9.4
4.3
1.3
61
22.3
10.2
3.1
112
28.4
18.7
5.7
204
44.6
34.1
10, a
173
58.1
28.9
8.9
599
30.7
TAMPERED
30.
81
32.0
21.9
4.1
90
32.6
24.3
4.6
61
22.3
16.5
3.1
70
17.7
18.9
3.6
46
10.1
12.4
2.4
22
7.4
5.9
1.1
370
16.9
C R 0 3 Z
ARGUABLE
31.1
141
55.7
14.9
7.2
156
56.5
16.5
8.0
146
53.3
15.4
7.5
209
52.9
22.1
10.7
201
44.0
21.3
10.3
92
30.9
9.7
4.7
945
48.4
i T A B U
MALFNCTr
0
33.
8
3.2
20.5
0.4
4
1.4
10.3
0.2
6
2.2
15.4
0.3
4
1.0
10.3
0.2
6
1.3
15.4
0.3
11
3.7
28.2
0.6
39
2.0
BY STATUS VEHICLE TAMPER STATUS
ROW
TOTAL
253
13.0
276
14.1
'
274
14.0
395
20.2
457
23. H
'
'. 298
[ 15.3
[
[
j
1953
100.0
CHI SUUARE
319.75342 WITH IS DEGREES OF FREEDOM SIGNIFICANCE & 0.0000
-------
Appendix V-2
-72-
-------
TESTING OF DATA ,AND PRQGHAM
FILE NONAME (CREATION DATE = 09/25/78)
09/25/78
PAGE 46*
STATt
COUNT 1
ROW PCT ]
COL PCT
TOT PCT
wA
VA
1
•-J •
^J TX
1
TN
ME
DE
COLUMN
TOTAL
STATUS
OK
0.
124
40.5
20.7
6.3
139
33. 1
23.2
7.1
36
16.7
b.O
1.8
111
3U.3
lb.5
5.7
103
30.7
17.2
5.3
86
2Q.2
14.4
4.4
599
30.7
TAMPERED
30.1
46 1
15.7 ]
13.0
2.5
59
14.2
15.9
3.0
48
22.2
13.0
2.5
66
20.4
17.8
3.4
70
20.9
18.9
3.6
79
22.2
21.4
4.0
370
18.9
CROSS
ARGUABLE
31.1
128
[ 41.8
13.5
6.6
207
49. B
21.9
10.6
129
59.7
13.7
6.6
144
44.4
15.2
7.4
155
46.3
16.4
7.9
182
51.1
19.3
9.3
9«5
48.4
1 T A B U
MALFNCTN
D
33.
6
2.0
15. tt
0.3
11
2.6
28.2
0.6
3
1.4
7.7
0.2
3
0.9
7.7
0.2
7
2.1
17.9
0.4
9
2.5
23.1
0.5
39
2.0
BY STATUS VEHICLE TAMPER STATUS
ROM
TOTAL
306
15.7
416
21.3
216
11.1
324
16.6
335
17.2
356
18.2
1953
100,0
CHI SQUARE B 55.89305 WITH 15 DEGREES OF FREEDOM SIGNIFICANCE = 0.0000
-------
Appendix V-3
-74-
-------
TESTING OF DATA AMI) PROGRAM
FILE NONAME (CREATION DATE
09/25/78
P^GE 162
09/25/78)
ui
I
MAKE
COUNT ]
RO-1 PCT ]
COL PCT
TOT PCT
VW
VOLV
TRIU
TOYO
SUBA
SAAB
RENA
m
COLUMN
TOTAL
STATUS
:
[OK
0.
13
35.1
2.2
0.7-
10
90.9
1.7
0.5
0
0.0
0.0
0.0
9
10.6
1.5
0.5
0
0.0
0.0
0.0
1
100.0
0.2
0.1
0
0.0
0.0
0.0 ]
599
30.7
TAMPERED
30.1
3 ]
8.1 1
0.8 1
0.2 I
1 J
9.1 ]
0.3 3
0.1 1
0 1
0.0 ]
0.0 ]
0.0
4
1.7
1.1
0.2
0
0.0
0.0
0.0
0
0.0
0.0
0.0
0
o.o
0.0
[ 0.0
370
18.9
C R 0 S 5
ARGUABLE
31.
18
48.6
1.9
0.9
0
[ 0.0
[ 0.0
0.0
3
[ 100.0
[ 0.3
0.2
72
84,7
7.6
3.7
8
100.0
0.8
0.4
0
0.0
0.0
0.0
1
100.0
0.1
0.1
945
48.4
J T A B U
MALFNCTN
0
33.
3
8.1
7.7
0.2
0
0.0
0.0
0.0
0
0.0
0.0
0.0
0
0.0
0.0
0.0
0
0.0
o.o •
0.0
0
o.o
0.0
0.0
0
0.0
0.0
0.0
39
2.0
BY STATUS VEHICLE TAMPER STATUS
ROM
TOTAL
37
1.9
11
0.6
3
0.2
85
4.4
8
0.4
1
0,1
1
0.1
1953
100.0
(CONTINUED)
-------
TESTING OF DATA AND PROGRAM
FILE NONAME (CREATION DATE
09/25/78
PAGE 163
09/25/78)
MAKE
COUNT 1
ROW PCT 1
COL PCT
TOT PCT
PONS
PONT
1
f PLYM
I
PEUG
OPEL
OLDS
MG
COLUMN
TOTAL
STATUS
OK
0.
1
50.0
0.2
0.1 '
47
35.3
7.8
2.4
26
24.5
4.3
1.3
0
0.0
0.0
0.0
I 1
I 100.0
0.2
0.1
52
36.9
8.7
2.7
0
0.0
0.0
0.0
599
30.7
TAMPERED
30.1
1 I
50.0 1
0.3 1
0.1 1
31
23.3
8.4
1.6
27
25.5
7.3
1.4
0
0.0
0.0
0.0
0
o.o
0.0
0.0
21
14.9
5.7
1.1
2
33.3
0.5
0.1
370
18.9
C H 0 S S
ARGUABLE
31.
0
0.0
0.0
0.0
S3
39.8
5.6
2.7
51
46.1
5.4
3
100.0
0.3
0.?.
0
0.0
0.0
0.0
66
46.8
7.0
3.4
4
66,7
0.4
0.2
945
48.4
TABU
MALFNCTN
D
33,
0
0.0
0.0
0.0
2
1.5
5.1
0.1
2
1.9
5.1
0.1
0
0.0
0.0
0.0
0
0.0
0.0
I 0.0
[ 2
1.4
5.1
0.1
0
0.0
0.0
0.0
39
2.0
BY STATUS VEHICLE TAMPER STATUS
ROM
TOTAL
2
0.1
133
6.8
106
5.4
3
0.2
1
0.1
141
7.2
6
0.3
[
[
1
1953
100,0
(CONTINUED)
-------
TESTING OF DATA -AND PROGRAM
FILE NONAME (CREATION DATE s 09/25/78)
09/25/78
PAGE 164'
MAKE
COUNT I
RQrt PCT 1
COL PCT
TOT PCT
MERS
MERC
MCOS
1
-J
-J
MAZD
MAZA
LINC
LANC
COLUMN
TOTAL
STATUS
OK
o.
1
100.0
0.2
0.1
20
26,0
5.3
1.0
I
100,0
0,2
0,1
3
27,3
0,5
0,2
1
100,0
0,2
0.1
8
38.1
1,3
o.«
1
100,0
0.2
0.1
599
30.7
TAMPERED
30.
0
0.0
0.0
0.0
21
27.3
5.7
1.1
0
0.0
0.0
0.0
0
0.0
0.0
0,0
0
0.0
0,0
0.0
3
1^.3
0.6
0.2
0
0.0
0.0
0.0
370
18,9
CROSS
ARGUABLE
31.
0
0.0
0.0
0.0
31
04.2
3.6
1.7
0
0.0
0.0
0.0
6
72.7 ]
O.A
0.4
0
0.0
0.0
0.0
9
12.9
1.0
0.5
• 0
0.0
I 0.0
I 0.0
945
08.4
TABU
MALFNCTN
D
33.1
0 ]
0.0
0.0
0.0
2
2.6
5.1
0.1
0
0.0
0.0
0.0
0
[ 0.0
I 0.0
[ 0.0
0
0.0
0.0
0.0
1
4.8
2.6
0.1
0
0.0
0.0
0.0
39
2.0
BY STATUS VEHICLE TAMPER STATUS
ROW
TOTAL-
1
0.1
77
3.9
1
0.1
11
0.6
1
0.1
21
1.1
1
0.1
1953
100,0
(CONTINUED)
-------
TESTING OF DATA AND PROGRAM
FILE NONAME (CREATION DATE • 09/25/76)
09/25/78
PAGE US
MAKE
COUNT 1
RO* PCT ]
COL PCT
TOT PCT
JEEP
INTE
MONO
]
i 1
£
I_
•
CMC
FORD
FIAT
DOOG
COLUMN
TOTAL
STATUS
OK
0.
2
50.0
0.3
o.i •
0
0.0
0.0
0.0
11
t 39.3
[ 1.8
[ 0.6
2
18.2
0.3
0.1
90
22.9
15.0
4.6
0
0.0
0.0
0.0
25
28.0
4.2
1.3
599
30.7
TAMPERED
30.
0
0.0
0.0
0.0
1
50.0
0.3
0.1
2
7.1
0.5
0.1
3
27.3
0.8
0.2
75
19.1
20.3
3.8
3
75.0
0.8
0.2
14
15.9
3.8
0.7
370
18.9
C R 0 3 {
ARGUABLE
31.
2
50.0
0.2
O.I
1
50.0
Oil
0.1
15
53.6
1.6
0.8
6
54. b
0.6
0.3
222
56. 5
2J.5
ll.o
1
25.0
0.1
0.1
48
54.5
5.1
2.5 ]
945
48.4
.TABU
MALFNCTN
D
33.
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
0.0
6
1.5
15.4
0.3
0
0.0
0.0
0.0
1
[ 1.1
[ 2.6
C 0.1
39
2.0
BY STATUS VEHICLE TAMPER STATUS
ROM
TOTAL
4
0.2
2
0.1
28
1.4
11
0.6
393
20.1
4
0.2
88
4.5
1953
100.0
(CONTINUED)
-------
TESTING OF DATA•AMD PROGRAM
PILE NONAME (CREATION DATE a 09/25/78)
09/25/78
PACE 1664
MAKE
COUNT ]
ROW PCT
COL PCT
TOT PCT
OATS
CHRY
CHEV
1
-J
X>
1 °
CADI
BUIC
BMW
AUDI
COLUMN
TOTAL
STATUS
[
[OK
0.
35
61.4
5,6
1.8 *
Ib
31.3
2.S
0.8
144
32.1
24.0
7.4
27
55.1
41
41.0
6.8
2.1
3
100.0
o.s
0.2
1
14.3
0.2
0.1
599
30.7
TAMPERED
30.
6
10.5
1.6
0.3
9
18.8
2.4
0.5
88
19.6
23.8
4.5
5
10.2
1.4
0,3
29
29.0
7.8
1.5
0
0.0
0.0
0.0
1
14.3
0.3
0.1
370
16.9
C R 0 S 5
ARGUABLE
31.1
14 J
24.6
1.5
0.7
23
47.9
2.4
1.2
203
«5.3
21.5
10.4
16
32.7
1.7
o.a
29
29.0
3.1
1.5
0
0.0
0.0
0.0
3
42.9 ;
0.3
0.2
945
46.4
5 T A B U
MALFNCTN
D
I 33.
[ 2
3.5
5.1
0.1
1
2.1
2.6
0.1
13
2.9
33.3
0.7
1
2.0
2.6
0.1
[ 1
[ 1.0
[ 2.6
E 0.1
t 0
[ 0.0
[ 0.0
[ 0.0
[ 2
[ 28.6
t 5.1
[ 0.1
39
2.0
BY STATUS VEHICLE TAMPER STATUS
ROM
TOTAL
57
2.9
48
2.5
446
22.9
49
2.5
100
5.1
3
0.2.
7
0.4
1953
100.0
(CONTINUED)
-------
TESTING OF DATA AND PROGRAM
FILE NONAME (CREATION DATE B 09/35/78)
09/25/78
PAGE 167
MAKE
STATUS
COUNT I
ROW PCT IOK
COL PCT I
TOT PCT 0.
AMC
COLUMN
TOTAL.
13
1
0
8
.3
.3
TAMPERED
30.
33
S
1
20
.3
.4
.0
0 S
ARGUABLE
31.
3
1
32
!«
.6
599 370 945
30.7 18.9 48,4
S
I
I
I
I
I
TABU
MALFNCTN
D
33.
0
0.0
0.0
0.0
39
2.0
I
I
I
I
I
I
I
BY STATUS VEHICLE TAMPER STATUS
ROM
TOTAL
60
3.1
1953
100.0
CHI SQUARE a 267.76416 WITH 105 DEGREES OF FREEDOM SIGNIFICANCE s 0.0000
oo
o
I
-------
Appendix VI-1
-81-
-------
TESTING 0^ DATA AND PNUUHAM
FILE. NONAME (CHEAIIUN DATE = 09/26/70)
09/26/78
PAGt
22
CRITtKlUlN VARIABLE IHC
bKUKCN D04N BY LIMCAP
dY MYR
VAHIAbLE
FOR tNTiHt POPULATION
IDLE HC
LlMlTtR CAPS
CODE VALUE LABEL
OF S U b P
SUM
295109,0000
MEAN STO OtV VARIANCE
196,0857 259.4655 67322.3218
N
( 1505)
LIMCAK
MYK
MfH
MYN
MYK
MYK
MYK
LIMCAK
, MYK
30 MYK
••^ MYK
1 MYK
MYN
MYK
LIMCAH
MYH
MYK
MYK
MYK
MYK
MYK
TOTAL CASES *
MISSING CASES a
1953
448 OH
0.
73.
7«.
75.
76,
77.
12.
73.
7«.
VS.
76,
77.
78.
98.
73.
71.
75.
76.
77.
78.
22.9 PCT.
FUNCT PROPERLY
ARGUABLE TAMPERING
NOT EQUIPPED
56391,
5645.
4511.
5840.
12485.
20805.
7105.
0000
0000
0000
0000
0000
0000
0000
227108
51695
54038
20166
55959
31280
5970
11610
1190
3745
1850
2600
1615
610
.0000
.0000
.0000
.0000
,0000
,0000
.ouuo
.0000
.0000
.0000
.0000
.0000
.0000
.0000
116.5103
162.0966
180.4400
121.6667
145.1744
127.638U
54,2366
239.8164
29*.0977
296.9121
167.7733
259.0694
168.4337
101.1864
156.8919
236.0000
312.0633
142.3077
216.6667
76.9046
55.4545
163.7625
24J.26S2
211.9735
195.1931
160.2600
161.6020
90.2464
290.3218
342,6844
360.5927
225.2486
275.4513
212.9524
150.095t>
221.2878
171.0848
380.1403
207. 2098
209.2990
99.1019
77.0861
26818.1510
59 17 /.ViiVO
44932.7567
36100.3546
25689.6751
26115. 22UO
8144.41<:8
842B6.74.i9
117432. M973
130027. 11V3
50/36.9261
75U/3.4324
45346.7441
22528. 7405
48968.
-------
TESTING OF DAT* ANO PMOGRAM
CRITtrflUN VARIABLE IHC
VARIABLE
LIMCAH
LIMCAH
LIMCAP
CODE
0.
12.
09/26/70
P*GE 23
.... ANALYSIS
VALUE LAbEL
FUNCT PROPERLY
ARGUABLE TAMPERING
NOT fcUUIPPED
WITHIN GROUPS TOTAL
or v A K i
SUM
56491,0000
22/108,0000
11610,0000
MEAN
116.5103
239.6184
156,8919
STO DEV
163.7625
290.3218
221.2878
SUM OP SO
12953166.9403 (
79735258. /603 (
35/4685.1351 (
H
484)
94/)
74)
295109.0000
196.0857
253.1599 96263110.8438 (
1505)
I
oo
OJ
I
SOURCE
BETWEEN GROUPS
WITHIN GROUPS
ANALYSIS 0
SUM OF SUUARES
4989661,099
96263110.844
ETA s 0.2220
F V A
R I A N C £
D.F. MEAN SQUARE
2 **********
1502 64089.954
ETA SUUAREO s 0.0493
*
*
*
*
F SIG. *
*
38*927 0.0000 *
*
*
*
*
*
-------
Appendix VI-2
-34-
-------
TESTING OF DATA'ANU PWBGRAM
09/26/78
PA&E
FILE NUNAME
OATE = 09/26/70)
CNlrtKlUN VARIABLE 1CU
tfKUKtN DQrtN dY LIMCAP
8Y MYK
VARIABLE
FUR ENTIRE POPULATION
LIMCAK
MYK
MYK
MYK
MYK
MYK
MYK
LIMCAH
M1K
, MYK
00 MYK
U1 MYK
1 MYK
MYH
LIMCAP
MYK
MYK
MYK
MYK
MYK
MYK
IDLE CU
LIMITER CAPS
CODE VALUE LABEL
0. FUNCT PROPERLY
73.
75!
77!
78.
12. ARGUABLE TAMPERING
'3.
74.
75.
76.
77.
78.
98. NOT EQUIPPED
73.
74.
75.
7<».
77.
78.
OF 5 u a P 0
SUM
3945.3997
633.U999
87.2000
70.6000
57.3000
147.2000
173.4000
98.0000
3136.6998
712.0000
764,8999
420.0000
665.0999
444.1000
130.6000
174.8000
25.0000
36.7000
25.6000
46.3000
28.4000
12.8000
PULAIIUNd
MEAN
2.0243
0.9920
2.3568
2.2125
0.98/9
1.3505
0.8065
0.5213
2.6117
3*4067
3.3548
2.18/5
2.5193
2.0656
1.4043
1.6037
4.1667
2.4467
1.0667
2.2048
1.0519
0.8000
STO DEV
2.4761
1.7758
2.3781
2.5494
1.5871
1.9M50
1.5411
1.3679
2.6355
2.7361
2.7872
2.5997
2.5294
2.3386
2.0227
2.1110
3.4679
2.4b9<4
1.6972
2.4622
1.1630
1.3317
VARIANCE
6.1310
3.1536
5.0553
6,4992
2.5190
3 . 9 <1 U 1
2.3750
1.8711
6.9461
7.4861
7. 7685
6./5B3
6.3979
ii ii /i y i
n U O 1 X
n fi ^ fa "I
12iu2o7
6.19/0
2. 8806
6.1615
1.3526
1./733
N
( 19«9)
( 639)
( 37)
( 32)
( 5d)
( 109)
( 215)
( 188)
( 1201)
( 209)
( 228)
( 192)
( 2b4)
( 215)
( 93)
( 109)
( 6)
( IS)
( 24)
( 21)
( 27)
( 16)
TOTAL CASES =
MISSING CAStS a
1953
4 OK 0.2 PCT.
-------
TESTING UK DATA AND PHQGHAM
CKlTtKjON VARIABLE ICU
09/26/78
PAGE 25
VARIABLE
L1MCAI*
LIMCAP
LIMCAP
oo
CODE VALUt LAbEL
0. FUNCT PKUPEHLY
12. ARGUABLE TAMPERING
96. NOT EUUIPPED
WITHIN GROUPS TOTAL
ANALYSIS 0
SOURCE SUM OF SUUARES
BETWEEN GROUPS 1114.655
WITHIN GROUPS 10620.571
ETA c 0.3055
a i a u r VAKI
SUM
633.8999
3136.6998
171,6000
3945.3997
F VARIANCE
D.F. MEAN SQUARE
2 557.327
1946 5.565
ETA SQUARED = 0.0933
MEAN STD DEV
0.9920 1.775H
2.6117 2.6355
1.6037 2.1110
2.0243 2.3589
*
*
*
*
F SIG. *
*
100,157 0.0000 *
*
*
SUM Of- SO N
2011. 96V1 ( 639)
8335.3239 ( 1201)
481.<»7tJ5 ( '109)
10628.5714 ( 1949)
-------
Appendix VI-3
-87-
-------
FILt
Ut UAfA AMU
NUMAME (CKtAllUN UATfc = 09/26/78)
09/26/78
PAGE
IB
CKlTtKlUN VAHlAHLt IMC
UKUKtN DO-IN dY tGKSl
6Y MYK
VAHlAbLt
FOK tMiHt POPULATION
EGHSUT
MYK
MYK
Mf K
MYK
MYK
EGHSTftF
MVr<
MYK
MYK
MYK
MYK
MYK
EtrtSTAT
MTK
MYH
MYK
MYK
MYK
MYK
EbHSUf
MYK
MYK
MYK
MYH
MYK
MYK
EUKSfAT
MYK
--•- OE3CMIPTION
IOLE HC
AT EliK SYSltM STATUS
CODE VALUE LABEL
0. OK
73.
yij
76J
77.
78.
30. TAMHtKEO
73.
74.'
Tb\
77.
78.
33. MALhNCTNO
73!
75.
/6,
77.
78.
98, NUT tQUIPPtO
73.
7«jt
/6.
77.
78,
9V, NON STOCK EQUIH
73,
OF 3 U 6 H
SUM
295109.0000
191546.0000
21517.0000
33382. oOC'O
21220 . OOUU
56962 ,0000
43615.0000
11850. 0000
498UO.OOOO
14951.0000
14//2.000U
6591.0000
8276.0000
49 40. 0000
360.000U
19256.0000
9960.0000
4735.0000
1255.0000
2811.0000
460. UOOO
15.0000
32051.0000
9706.0000
9405.0000
3/90.0000
2995.0000
4695.0000
1460.0000
23/6.0000
.2376,0000
OPULATIUN
MEAN
196. 085/
181.56U2
247.3218
295. U 159
163.6UU6
235,3802
151.9686
66.5730
266.7380
347.6977
321.1304
205.96B8
236, u5 /I
205.1 167
51,4286
200.5833
219.5000
109.4000
1 S9.4444
200. 7b5/
9^,0000
5.DOOO
194.24H5"
255.421 1
2hU, / 143
1/2.2/^7
130.2174
13a.08H2
112.30/7
1188.0000
1188.0000
STO DEV
259.4655
242.5808
296.5500
351.6840
212.4025
264.0120
183,9353
114.1713
329.77B7
390.6443
421.2063
226.1573
231.0092
249.4686
37.2731
195.1446
232.0665
153.55/3
161.13/4
iao.4551
59.3296
8.6603
258.2238
292.7691
3i'4.2603
2/2.0457
148,2U17
185.0963
115.8234
1149.7556
1149,7556
VARIANCE
67322.3218 (
58845.4269 (
87941. B952 (
123f.bl.63HO (
4511'i.tt2
-------
TESTING Uh DATA «NO PKUUHAM
CHIUNIUN VARIABLE IHC
OV/46/7B
19 ,
VARIABLE CUDt VALUE LABEL
EGH3IAT 0. UK
tliKSlAT 30. TAMHtHEO
tChSUT 33. MALFNCTND
fcGHSUT 98. NOT EulJIPPED
EkKSTAT 99. NON STUCK fcUUIP
WITHIN GHOUP3 TUTAL
ANALYSIS U
SOUHCE SUM OF SUUANtS
HETrttEN (jKUUPS 3126337.709
WITHIN GHOUP3 Vttl(26iJ34,2J4
tTA a 0.17*7
SUM MEAN STU DEV SUM UF SU N
1V1546.0000 1BI.S602 2')2.bHOB 6202307^.9200 ( lObb)
U9080.0UOO 266.73MO 329.77tt7 2022h240, 1604 ( 187;
320'3l!ouOO 194^248^ 2^8.2238 10935442.8121 ( 16SJ
23/6.0000 11H8.0000 1149.7556 132193d. 0000 C 2)
2V5109.0000 196.0057 21i5.7687 98126434.2338 { IbOS)
F VARIANCE
D.F, MEAN SQUARE F -SIC,
4 781b84,427 11,948 0.0000
liOO 6S017.623
tTA SUUArfED s 0.0309
-89-
-------
Appendix VI-4
-90-
-------
TESTING U)- UATA AMU PHUteHAM
FlLt NUMAME (CHtAIIUN DATE a 09/26/76)
09/26/78
20 „
CHlltMiUN VAHlAhLt ICU
riKUftLN DUNN dY tbHSlAT
dY MYH
VAHlAOLt
FUH tNMKt POPULATION
EGRSTAT
MYH
MYH
MYK
MYH
MYK
EGKS1AT
MYK
MfH
M f k
n i n
MYH
M f K
MYH
tUKSIAT
M YH
ni«i
MYH
MYK
MYK
MYH
MYK
fcGKSUT
MYK
MYK
MYK
MYH
MYK
MYK
MYH
- - - UESCK1PTIUN
IDLfc CU
LGH SYSlfcM STATUS
COOt VALUE LABEL
0, OK
73,
74,
7b.
76,
77
' ' ,
30, TAMPtKfcU
74!'
7«j,
76,
77,
33, MALf-NCTNU
74,
7b.
77,
78,
9B, NOT E&UIPPED
73.
• -* f
7 il
7b,
76,
77,
78,
99. NUN STOCK LUUIH
OF 3 U B P
SUM
394S.399/
2«b2.;-J9H
334.7000
441,bUOU
314,'luOO
529!bOOO
ib/.ouoo
bUb.OOOO
21d.'4UOll
202.2UOO
b2,-4UOO
123.2000
4 tt , *y o o o
10.9000
3//.6000
Ibb. luOO
lOb.bOUO
bl ,2oOO
34.9UOO
li.bUOO
lb. 200(1
406.4UOU
1U3,4UUO
122.0000
b4.9uOO
44.9UOO
53.9000
27,3000
12,6000
12,6000
UPULAT IONS
MEAN
2,0243
l,78Hb
3.0427
3,0667
1./2/5
2.1926
1.39/1
0,7110
3,0901
3|h/b4
2,1 bt34
3,0049
1 , 74<>4
1.2111
2.81/9
3.3717
2.9bll
2,bbOO
1 ,H3b8
1,9429
2.7000
1,8991
2,404/
3.0bUU
1.6147
1.2829
1 .2b35
1.43b8
6,3000
6,3000
STO UEV
2,47fal
2i684H
2.d23/
2.430b
2'.0704
I,b238
2,7842
3,0067
2./913
2 . 2248
2.76bb
2.2006
1.33b5
2.6/29
2.bilO
2.5314
3.3343
2.2234
1 ,9dUb
3.1b43
2,1416
2.193b
2.74i9
1 . 7245
1.5240
1.5956
2.2993
0.4243
0.4243
VAklANLt
6.1J10 (
5.7047 (
7)9 1 i4 (
5.9077 (
6.0557 (
4.2bb7 (
2.3220 (
/./Sltt (
7!791b (
4 , 94 Vb (
7,b4t>0 (
4 , H4f*6 (
l./bi6 (
7.14
-------
TESTING Uh DATA AND PROGRAM
CNITtHlUN VARIABLE ICU
09/26/78
PAGE 21
VAH1AULE CUDt VALUt LAUEL
tbMSIAT 0, UK
EGH6TAT 30. TAMPERED
EGH8TAT 33. MALFNC1NU
EGHSUT 98. NOT tUUlHPEO
EGHSUT 99. NON STOCK EQUIP
WITHIN GKUUPS TUTAL
ANALYSIS 0
SUUHCE SUM OF SQUANtS
BETWEEN liRUUKS 4b3.033
MIThIN GKUUPS 11490.193
t[A a 0.1V48
sis a ^ v A H i
SUM
6H6.0000
3//.6000
t06.4000
12.6000
3945.3997
f VARIANCE
O.F. MEAN SlJUAHt
4 11>.2S8
1944 b.911
ETA SQUARED * 0.0379
MEAN STU OEV SUM Uf SU N
1,/bttS 2.3«85 7tt49./HJO C 1377)
4.0901 2./042 1713. 13ol ( 222)
2.B1/9 2.6729 950.2369 ( 134)
1.8991 2.1416 976.91V7 ( 214)
6,3000 0.4243 0.1800 ( 2)
2.0243 2.4312 11490.1928 ( l"»49)
F SIG.
19.162 0.0000
-92-
-------
Appendix VI-5
-93-
-------
TESTING OF DATA AND PKUtJRAM
ULE NUNAMt (CREATION DATt = 09/26/78)
09/26/78
PAGE
CKITtKlUN VAKIAULE 1HC IDLt HC
uKUKtN ooviH at &TATUS VEHICLE TAMPER STATUS
Q V U V U
O T " T «
VAKIAbLb
FOR tNTIHE POPULATION
STA!US>
MYK
MYK
MYK
MYK
MYK
MYK
STATUS
MYK
MYK
MYK
MYK
MYK
MYK
STATUb
MYK
MYK
MYK
MYK
MYK
MYK
STATUS
MYK
MYK
MYK
MYK
MYK
MYK
CODE VALUE LAUEL
0. OK
73.
74.
75.
76.
77.
78.
30. TAMPERED
73.
74.
75.
76.
77.
78.
31. ARGUABLE
73.
74.
75.
76.
77,
78.
33. MALFNCTND
73.
74.
75.
/6.
77.
78.
SUM
295109.0000
47266.0000
2350,0000
4606.0000
5060.0000
12385.0000
16200.0000
6645.0000
79569.0000
22637.0000
22532.0000
101 16,0000
14744.0000
8620.0000
920.0000
163359.0000
32723.0000
33126.0000
20520.0000
42865. OOOu
28530.0000
5595.0000
4915,0000
820.0000
2030.0000
140.0000
1050.0000
350.0000
525.0000
MEAN
196,0857
109.1594
123.6842
230,3000
115.4545
151.0366
108.7248
55.8403
261.7401
32*. 0725
312.9444
210.7500
254,2069
221,0256
51.1111
222.2571
282.0948
269.31 71
180.0000
252.1471
162.8846
99,9107
148,9394
136.6667
507.5000
28.0000
262.50UO
58.3333
65.6250
STD DEV
259.4655
155.9874
123,1738
295.1506
180.2770
166.1078
145.8551
93.0129
311.3740
368.4803
382.4806
204.6237
221.9911
252.2691
51.5796
272.9193
312.6801
333.9081
236.6011
287.5100
198.3798
153.5415
218.9027
68.0152
417.2429
40.8656
212.1910
49,1596
102,7264
VARIANCE N
67322.3218 ( 1505
24332.0602 ( 433
15171. /83t> ( 19
87113.9053 ( 20
32499.7866 ( 44
27591,6135 ( 62
21273.7008 ( 149
6651.4065 ( 119
96953.7771 ( 304
150916. 9211 ( oV
146291,^772 ( 72
41952.7447 ( 43
49260, 0617 ( 58
63639,7099 ( 39
2660.4575 ( 18
74484.9188 ( 735
97/66.6170 ( lit*
111494. t.2dl ( 123
56074.7768 ( 114
82662.0197 ( 170
39354.5285 ( 156
23574.9919 ( 56
47918.3712 ( 33
7746.6067 ( 6
174091,6667 ( 4
1670.0000 ( 5
45025. 0000 ( 4
2416.0667 ( 6
10553.1250 ( 8
TOTAL CASES =
MISSING CASES 8
448 UK 22.9 PCT,
-94-
-------
TESTING OP DATA AND PROGRAM
CKIUK1UN VARIABLE IHC
09/26/78
PAGE
VARIAbLE CUD£ VALUfc LABEL
STATUS 0. OK
STATUS 30. TAMPtRED
STATUS 31. ARGUABLE.
STATUS 33. MALFNCTMD
WITHIN GROUPS TOTAL
ANALYSIS 0
SOURCE SUM OF SQUARtS
BETWEEN URUUPS 5159009.189
WITHIN GROUPS 9609 3762. 754
ETA c 0.2257
a I a u r VAKI
SUM
47266.0000
7*569,0000
163359,0000
4915.0000
295109.0000
F VARIANCE
O.F. MEAN SUUARE
3 **********
1501 64019,829
ETA SQUARED = 0.0510
MEAN STD OEV SUM OK SO N
109.1594 155.9874 10511450.0046 ( 433)
261.7401 311.3740 2937699(1. 4704 ( 304)
221.2571 2/2.9193 546M93G . 40UO ( 735)
148.9394 218.9027 1533387. b7H8 ( 33)
196.0857 253.0214 96093762.7538 ( 1505)
F S1G,
2 6'. 862 0.0000
-95-
-------
Appendix VI-6
-96-
-------
TESTING OF DATA AMU PWOGHAM
FILE NUNftME (CHfeAFIUN DATE = 09/26/78)
09/26/70
PAGE
CKITtKlUN VARIABLE KU
6KOKE.N DOrtN 6Y STATUS
8Y MYR
VAHIAbLE
FOR ENTIRE POPULATION
STATUS
MYK
MYK
MYK
MYK
MYK
MYK
STATUS
MYK
MYK
MYK
MYK
MYK
MYK
STATUS
MYK
MYK
MYK
MYK
MYK
MYK
STATUS
MYK
MYK
MYK
MYK
MYK
MYK
IDLE CO
VEHICLE TAMPER STATUS
CODE VALUE LABEL
0. OK
73,
74,
75.
76.
77.
78,
30, TAMPERED
73,
74.
75.
76.
77.
78,
31. ARGUABLE
73.
74.
75.
76,
77.
78,
33. MALFNCTND
73,
74,
75.
76,
77,
78.
OF S U B P
SUM
3945.3997
562,3999
44.0000
68.4000
65,bOOO
160.5000
140.2000
63.0000
1057.5999
315,9000
308.90UO
151.8000
186. SOUU
81.5000
13.0000
2259,8996
443.7000
4B2.0000
284.5000
503.5000
421.2000
125.0000
65.5000
19,8000
13.1000
1.1000
8.1000
3.0000
20.4000
OPULATIONS
MEAN
2.0243
0,9405
1.9478
2.6308
1,0738
1.4330
0.6873
0,4tt26
2.U661
J.9487
3.4322
2.4885
2.6643
1.7717
0,5909
2.3965
3.1468
3.1097
1.9406
2.4207
2,0955
1,3587
1.6795
2.47bO
3.2750
0.1833
2,0250
0.5000
1.8545
STO DEV
2.4761
1.6939
2.2078
2.7727
1.5467
2.0U39
1.3565
1,2278
2.7055
2,9465
2.6931
2.6204
2.50/9
2.0555
1,0479
2.5726
2.6111
2.7975
2.5519
2.5435
2,3550
2.0363
2.3095
2.0197
3.6234
0.4021
1.7251
0.7899
2.8991
VARIANCE
6.1310 (
2.8694 (
4,6744 (
7.6B78 (
2.3923 (
4.1777 (
1.6401 (
l.bU/5 (
7,3197 (
8.64J1B (
7,25cib (
6.6667 (
6.2696 (
4.22b2 (
1,09150 (
6.6185 (
6 . B 1 7 9 (
7.626
-------
TtSTlNG OF DATA AND PKOGKAM
CHITtKiUN VARIABLE ICO
09/26/78
PAGE
VARIABLE
STATUS
STATUS)
STATUS
STATUS
i
1
CODE VALUt LAbEL
0. OK
30. TAMPtKED
31. ARGUABLE
33. MALFNCTND
WITHIN GROUPS TOTAL
ANALYSIS 0
SOUHCE SUM OF SQUARES
BETWEEN GROUPS 1099.238
WITHIN GROUPS 10843.988
ETA = 0.3034
*
SIS U r V A K I «
SUM
562.3999
1057.5999
2259.8998
65.5000
3945.3997
F VARIANCE
O.F. MEAN SQUARE
3 366.413
1945 5.575
ETA SQUARED = 0.0920
MEAN
0.9405
2,8661
2.3965
1.6795
2.0243
F SIG.
65.721 0.0000
STO DEV
1.6939
2.7055
2.5/26
2.3095
2.3612
*
*
*
*
ft
*
*
SUM OK SO N
1713.0605 ( 59b)
2693.6463 ( 369)
6234.5960 C 943)
202.oB.i6 ( 39)
10843.9664 ( 1949)
-98-
-------
Appendix VI-7
-99-
-------
TEST lNli Uh DATA AMU PKOGHAM
FILE NUNAMfc. (CHEATIUN UATE a 10/26/78)
10/26/70
PAGE
PbSTAI
LEAD CUNC STATUS
CHOSSFABULATIUN OF ********
BY FILSTAT FULEH NECK STATUS
PBSTAT
PASS
FAIL
o
o
PAGfc 1 UF 1
FILSTAT
COUNT I
NO* PCT IPASS FAIL
COL PCT 1
TOT PCT i o.i i.
0. I
I
I
I
1. I
I
I
I
_ f _
•*•
COLUMN
TOTAL
455
98.3
96.0
9 a ,'2
IS
7b.O
J.2
3.1
e
1.7
61.5
1.7
5
25.0
38,5
1.0
U6U 13
97.3 2.7
HUH
TUTAL
461
9S.B
20
a. 2
481
100.0
COKRECTtO CHI SQUAKE 8 31.10187 KITH 1 DEGREE OF FREEDOM SIGNIFICANCE » 0.0000
-------
|