US Environmental Protection Agency
Office of Pesticide Programs
The Pesticide Handlers Exposure Database
Reference Manual
February 1995
-------
PHED: THE PESTICIDE HANDLERS EXPOSURE DATABASE
REFERENCE MANUAL
Version 1.1
Prepared by:
Versar Inc.
6850 Versar Center
Springfield, VA 22151
Prepared for
The PHED Task Force
representing
Health Canada
The U.S. Environmental Protection Agency
American Crop Protection Association
February 1995
-------
-------
FOREWORD
The Pesticide Handlers Exposure Database fPHED^ is a a»*~u • •
d"1 f0r iovolvcd fa "* or awlicaion^SLtoTto
field. This database was riwignH hv a Task Fom#» -* •
Camda (HC), to U.S. £2ZSJ *?" *?*
Protection Association (ACPA). The software was developed by Venar
consulting firm located in Springfield, Virginia. environmental
The basic assumption underlying the database is that exposure to pesticide users i,
primarily a function of the physical parameters of handling andlpplication
type and mixing/loading/application procedures), rather than 0f the rh»m^i nronertiM of th
active ingredient. IHe Task Force feels that ihere wiHe2£
5^ ^°W exposm. ^ ^ assessments to be conducted with a much larger degree
of certainty, since exposure estimates will be based on a much larger n.,n,w VaE.
^f^,ft0margIe"?OSraeStUdy-
tool for determining and comparing the influences of different parameters on pesticide exoosures
^eglsttants wiI1 be able to use the database to make risk fnan^n,^ decisions
at die product design stage, substantially prior to the registration stage. Lastly the cost of
ixstrade exposure evaluations will be reduced through the better usTof resour£s withh?tte
sooitific community to fill data gaps apparent from the uneven distribution of records in PHED
across formulation types and application methods. rahn
The current version of PHED (VI. 1) is an incremental improvement over the 1 m
version. With the addition of significantly more field studies, pesticide application methods and
studies using whole-body dosimeters, the user can analy^a greateV wte of
PHEDnot^S ±C database still has some limitations and it should be cautioned that
PHED not be considered a panacea in estimating pesticide handler exposure Accurate
subsetong and interpretation of exposure scenarios and values are essential in the proper
utilization of the data base. The following documents are included with each PHED pacteS
U^'s G^, Evaluation Guidance, Guid^ce for
Reportmg PHED Exposure Evaluations, and Data Entry Diskette User's Guide. TTie PHED
1S ®mwtly evaluating the requirements for a new version of PHED (V2.0)
scheduied for release in 1997. This version, which will be written as a stand-alone application'
win include a more user-friendly" interface and provide many new and improved fonctions.
-------
-------
TABLE OF CONTENTS
Page
1. INTRODUCTION ^
2. PHED USER'S GUIDANCE SECTION 2-1
2.1 Installation and Logging On 2-1
2.1.1' First Time Installation of PHED 2-1
2.1.2 Installation of PHED Updates 2-2
2.1.3 Logging Onto PHED 2-3
2.2 Menus 2-4
2.2.1 Definition of Menu 2-4
2.2.2 Execution of Menu Options 2-4
2.2.3 On-Screen Messages 2-6
2.2.4 Listing of PHED Menus . . . 2-6
2.3 Data Entry Screens . . . 2-7
2.3.1 Using Data Entry Screens 2-7
2.3.2 Entering Data 2-7
2.3.2.1 Data Entry Prompt 2-7
2.3.2.2 HELP Screens 2-10
2.3.2.3 Automatic Field Advance 2-10
2.3.2.4 Multiple Choice Data ...2-10
2.3.2.5 Multivalue Fields 2-12
2.3.2.6 Comments Field 2-13
2.3.2.7 "EDIT?" Prompt 2-14
2.3.3 Editing Saved Records 2-14
2.4 Data Analysis Screens 2-16
2.4.1 Creating Subsets 2-20
2.4.1.1 Subsetting Screens 2-20
2.4.1.2 Special keys 2-29
2.4.1.3 Steps in Subset Specification 2-31
2.4.2 Exposure Analysis Pathway 2-37
2.4.2.1 Inhalation Exposure 2-39
2.4.2.2 Dermal Exposure 2-41
2.4.3 Statistical Analysis Pathway 2-53
2.4.3.1 Correlation 2-61
2.4.3.2 Regression 2-62
2.4.3.3 Univariate Statistics 2-64
2.4.3.4 Statistics for "Complete" Data Sets Only 2-67
2.4.3.-5- ASCH File Creation 2-73
2.4.4 Browse/Print 2-73
2.4.5 Specifications 2-85
2.4.6 Delete 2-85
ii
-------
TABLE OF CONTENTS
Page
2.4.7 Print File/Subset 2-85
2.4.8 Print Comment 1 ! 1 1 1 1 2-90
2.4.9 List/Count 1 ! 1 1 1 1 2-90
2.5 Back Up/Restore of User Entered Data 2-90
3. STATISTICS DOCUMENTATION 3_j
3.1 Exposure Pathway 13-1
3.1.1 Inhalation Exposure Algorithms .....3*2
3.1.2 Dermal Exposure Algorithms ..................... 3.4
3.1.3 Frequency Distribution Testing 1! 3-8
3.1.4 Calculation of the Confidence limits for Total Body Exposure . . 3-13
3.2 Statistics Pathway
3.2.1 Technical Information 3_14
3.2.2 Non-Numeric Data 3_15
3.2.3 Normalization 111111 3-16
3.2.4 Accepting or Rejecting Inside Patch Value 3-16
3.2.5 Patch Variable Configuration .
3.2.6 Univariate Statistics " 3_j7
3.2.7 Correlation Statistics ] [ ] 3_17
3.2.8 Regression Statistics ] * [ 3.17
3.2.9 Performing Statistics on a Specified Variable 3-17
Appendix A - Data Entry Reference Tables ^-1
Appendix B - Exposure Survey Forms
111
-------
LIST OF TABLES
Eags
Table 1. Surface Areas for Regions of the Adult Body and
Locations of Dermal Exposure Pads that Represent These
Regions
Table 2. Matrix Listing Each Dermal Exposure Scenario and
the Applicable Patch Location Codes 3.5
Table 3. Procedure for Assigning Numeric Values to Exposure
Fields with Non-Numeric Data 3^
Table 4. Specific Body, Patch, and Surrogate Locations.Used
in the PHED Exposure Algorithms 3.7
Table 5. Surface Areas Used for Exposure Fstimatin^ 3_g
Table 6. Table of t-Statistics Values for. Testing Frequency
Distributions 3_12
Table 7. t-Statistics Table Utilized to Test for 95%
Significance of Correlation Results 3.19
iv
-------
-------
LIST OF FIGURES
Eass
Figure 1. Pesticide Handlers Exposure Database (PHED) System Structure 1-2
Figure 2. PHED Main Menu . 2-5
Figure 3. PHED Data Entry Menu 2-8
Figure 4. Example Data Entry Help Screen 2-11
Figure 5. PHED Data Analysis System Structure 2-17
Figure 6. Data Analysis: File/Subset Selection Morn 2-18
Figure 7. Subset Operations Menu 2-19
Figure 8. Subsetting Pathway 2-21
Figure 9. Subset Variable Selection 2-23
Figure 10. Subset Variable Specification 2-27
Figure 11. Example Coded Variable Selection 2-28
Figure 12. Example Subsetting Specification Screen 2-30
Figure 13. Exposure Algorithm Selection Menu 2-38
Figure 14. Example Inhalation Exposure Summary Statistics Screen 2-40
Figure IS. Dermal Exposure Scenario Specification Menu 2-43
Figure 16. Normal Work Clothing Menu 2-44
Figure 17. Protective Clothing Menu . 2-45
Figure 18. Head Patch Specification Menu 2-49
Figure 19. Example Dermal Exposure Summary Statistics Screen 2-50
Figure 20. Statistics Variable List Entry Screen 2-54
v
-------
LIST OF FIGURES (Continued)
Page
Figure 21. Patch Combination Selection Box 2-59
Figure 22. Inside Patch Selection Box 2-60
Figure 23. Example Correlation Screen 2-63
Figure 24. Example Regression Screen 2-65
Figure 25. Example Univariate Screen 2-66
Figure 26. Complete Data Set Criteria Selection 2-68
Figure 27. Example Total Body Univariate Screen 2-70
Figure 28. Example Total Body Correlation Screen 2-71
Figure 29. Pearson's Correlation Versus Spearman's Correlation 2-72
Figure 30. Browse/Print Variable Selection . . . 2-75
Figure 31. Browse/Print Variable Sort 2-79
Figure 32. Browse/Print Variable Selection List 2-81
Figure 33. Browse/Print Operations Menu 2-82
Figure 34. Example Browse/Print Display 2-84
Figure 35. Example Subset Specifications Display 2-86
Figure 36. Print File/Subset Menu 2-87
Figure 37. Example Matrix Format Printout 2-88
Figure 38. Example Survey Form Format Printout 2-89
*
Figure 39 Backup/Restore Menu 2-91 *
vi
-------
PESTICIDE HANDLERS EXPOSURE DATABASE REFERENCE MANUAL
1. INTRODUCTION
The Pesticide Handlers Exposure Database (PHED) is a menu-driven, user-friendly
generic pesticide exposure monitoring database designed and written by a Task Force
of representatives from Health Canada, the American Crop Protection Association (ACPA), and
the U.S. Environmental Protection Agency (EPA). The software was developed by Versar Inc.,
an environmental consulting firm located in Springfield, Virginia.
The environmental scientists and programmers at Versar used Revelation®, a **afa bay
management software package, to develop the series of data entry screens, data files, «nh«Awing
routines, exposure calculation algorithms, statistical analyses programs, print procedures, and
various supporting programs that comprise PHED. An overall system structure Hiagram is
shown in Figure 1.
The structure of this Reference Manual is as follows: Section 1 is the current
introduction; Section 2 is the PHED User's Guidance Section; and Section 3 and the Appendices
contain additional documentation for the system. Section 2 will be the most useful portion of
this Reference Manual for the average user. Section 2.1 describes the steps for installing and
logging onto PHED; Section 2.2 explains the basic structure and operation of the menus
contained in PHED; Section 2.3 describes how data may be entered into PHED, or edited;
Section 2.4 shows how to use the data analysis portion of PHED to creat subsets of data, to print
out summary exposure statistics, or to perform univariate, correlation, or regression statistics
on data in a subset; and Section 2.5 tells the user how to save and restore data entered by the
user when installing an update of PHED.
-------
£Q
S Ul
ll
<
a
m
O
If
gi
(L ¦
s
I
S'
i
*
1-2
jii
9 5
|5
n
s
3
U
s
55
E
• •
t
Q
111
S
2
2
Q
s
§
§.
iS
3
¦o
u
n
£
O)
E
it
112
i!
ii
el
o «
-------
2. PHED USER'S GUIDANCE SECTION
2.1 Installation and Logging On
Instructions are included in the Revelation* User's Manual for installing the Revelation*
software system on the user's personal computer. Revelation* and PHED require an IBM or
IBM-compatible personal computer system with at least (1) an AT or 386 capability; (2) 520K
RAM (random access memory); (3) MS-DOS Version 2.0 or later; and (4) one hard disk drive
and one 360K diskette drive. In addition, for speed of operation, it is suggested that the system
contains a math coprocessor.
2.1.1 First Time Installation of PHED
The Revelation* software system and the Pesticide Handlers Exposure Database (PHED)
must be installed on your personal computer before you log on to use PHED. If PHED is being
installed for the first time on your computer, complete the following steps in sequence:
1. Install Revelation* as directed in the Revelation* User's Guide. PHED WILL
OPERATE PROPERLY ONLY WITH REVELATION* G OR ITS RUN-TIME
VERSION. PHED WILL NOT OPERATE WITH ADVANCED REVELATION*
For more information on obtaining Revelation* G or its run-time version, consult
your local software dealer or contact ELF Software, 1419 Commerce Ave.,
Longview, Washington 98632, (800) 422-2511. -Any technical problems or
questions regarding Revelation* should be discussed with Revelation Technology,
Inc., at (206) 746-1629.
2. From DOS, change to the DOS subdirectory that contains the Revelation* system.
Usually this is C:/REV or D. /REV.
3. Ensure that the CAPS LOCK switch is on. During the PHED installation
process, Revelation* commands must be entered in uppercase.
4. Type REV SYSPROG to log on to the Revelation* SYSPROG account, followed
by the ENTER key.
2-1
-------
5. When the Revelation® log on menu appears, hold down the Ctrl key and press
the F5 key to access the terminal control language (TCL).
6. At the .• prompt, type CREATE-ACCOUNT PHED followed by the ENTER key.
7. When the password prompt appears, you may either press the ENTER key to
indicate that no password will be used, or you may specify a password. If you
specify a password, it must be entered each time you invoke PHED from DOS.
8. At the : prompt, type LOGTO PHED followed by ENTER key. If necessary,
enter the password you specified in step (7).
9. When the Revelation* log on menu appears, hold down the Ctrl key and press
the F5 key to access the TCL.
10. Place the PHED Installation Diskette 1 into your floppy dislr drive. At the :
prompt, type ATTACH A: followed by the ENTER key. If you use a drive other
than A, specify the correct drive.
11. At the : prompt, type RUN BP1 PHED.INSTALL followed by the ENTER key.
The information on the floppy diskette is being copied to the hard disk drive
while the drive light is on.
12. Follow the instructions that appear on the screen to load the remaining diskettes.
To cancel the installation process, press the ESCAPE key or the F9 key at any
prompt. When installation is complete, PHED will automatically begin to
operate.
2.1.2 Installation of PHED Updates
PHED updates, which may be necessary to-correct unanticipated system errors, will be
automatically provided to registered PHED users. If you have an earlier version of PHED on
your computer's hard disk drive and you wish to replace it with an update of PHED, complete
the following steps:
If you have entered any of your own survey records or have specified any subsets
from within the-main PHED files, be sure to back up this data before proceeding.
Your data can be backed up by selecting item 3 from the PHED Main Menu,
"Backup/Restore User-Entered Data". (See Section 2.5.)
2-2
-------
From DOS, change to the DOS subdirectory that contains the Revelation® system
Usually this is C:/REV or D:/REV.
Ensure that the CAPS LOCK switch is on. During the PHEn instaiiarir^
process, Revelation* commands must be entered in uppercase.
Type REV SYSPROG to log on to the Revelation® S YSPROG account, followed
by the ENTER key.
When the Revelation® log on menu appears, hold down the Ctrl key and press
the F5 key to access the TCL.
At the : prompt, type DELEIE-ACCOUNT PHED followed by the ENTER key.
A message explaining the two phases of deleting an account will appear. Press
the ENTER key to begin the deletion process.
At the prompt to begin phase 2 of the deletion process, type END followed by the
ENTER key.
Follow the installation instruction in paragraph 2.1.1 starting with step (6).
2.1.3 Logging Onto PHED
The Revelation® software (version G or its run time equivalent), the PHED program
software (a Revelation® application created by Versar, Inc.), and data files for PHED must
already be installed onto your computer's hard disk drive before you can log onto PHED. (See
Section 2.1.1.) The Log-On procedure is as follows:
1. From DOS, change to the DOS subdirectory that contains the Revelation® system.
Usually this is C:/REV or D:/REV.
2. Type PHED and press the ENTER key. This will invoke a batch file that was
automatically created during the PHED installation procedure. The Revelation
software will be loaded, and the PHED system will be invoked automatically.
3. If you specified a password during the PHED installation procedure, enter that
password in iijvpw-raw at the prompt
2.
3.
4.
5.
6.
7.
8.
9.
2-3
-------
4. The initial Revelation* screen will appear, and the measagf "Attaching for
the account PHED" will be displayed. After a moment, the initial PHED screen,
which is shown in Figure 2, will appear. Press any key to proceed to the PHED
Main Menu.
2.2 Menus
2.2.1 Definition of Menu
A menu is a screen listing the options that the user can select for the computer to
perform. Examine the sample menu in Figure 2, the "PHED Main Menu." Occupying the
center of the menu is the list of options the user can select for execution. These options may
bring up the data entry screens, execute access to other menus with additional options, generate
reports, or perform a number of other functions.
2.2.2 Execution of Menu Options
Menu items are selected by pressing the numeric key that corresponds to your choice.
You may either use the numeric keys across the top of your keyboard or, depending on your
keyboard, the numeric keypad on the keyboard's right-hand side. If you use the numeric
keypad, make sure that the NtiM LOCK switch is on. When a selection is made by either
method, the menu choice is highlighted to indicate that PHED has begun to process your choice.
Note; Unless then are specific instructions to the contrary (e.g., during processing of subsets,
exposure data, or statistics), you may cancel the current menu at any point in PHED and return
to the previous menu simply by pressing the F9 key or ESCAPE key.
2-4
-------
Pesticide Handlers Exposure Database
Main Menu
1. Data entry aenu
2. Statistical data analysis
3. Backup/restore user data
4. User-written option
Press the nunber of your selection
F9> Exit to DOS
FIGURE 2. ^hkii Main Miami
2-5
-------
2.2.3 On-Screen Messages
The top two lines of every menu screen describe the current menu and the current PHED
operation. The bottom two lines indicate which special keys (such as Function keys and the
ESCAPE key) are applicable within the current scr****"
2.2.4 Listing of PHED Menus
The following listing of major PHED menus is provided to give the user an overview of
menu names, menu functions, and how these menus interact with other functions of the database.
PHED Main Menu. This menu appears shortly after the user initially iogs on
to the PHED account; this menu provides the major pathway into the various
functions of the PHED software system.
Data Entry Menu. This menu accesses all the data entry screens for the PHED
software system. The four data files include the Applicator, Mixer/Loader
Flagger, and Mixer/Loader/Applicator (Combined) Files.
Data Analysis. This menu provides the user the opportunity to subset the data
within any of the four data files using the variables in the database to partition
data into the desired subset. The user may select a subset or data file to compute
exposures based on personal and protective clothing scenarios and/or to perform
various statistical analyses (univariate statistics, correlation, and/or regression).
Reserved for User Written Option. This space on the Main Menu is available
for users to write their own Revelation* programs for access via PHED.
Back Up/Restore User Entered Data. This option allows the user to back up
data that was not included with the original PHED data file. This option must be
used before updating the current version of PHED. After updating, the user must
use this option again to restore the data.
2-6
-------
2.3 Data Entry finrrnn
2.3.1 Using Data Entry Screens
Data entry screens are acceded by flra choosing a. Daa Entry option at the PHED
2 ZcTJ-D - C
2.3.2 Entering Data
Data entry screens are the means through which data are entered into this computer
ase. Appendix A contains reference tables for PHED data entry screens. The screens are
tata Arect.y entered onthe screen in the otder in which they occur on the Exp^ SuIVey
on^ ^ecu^r has been propammed to move fiom one daa entry field to the next in the
order of the fields on the Exposure Forms; i.e., when data are entered, or the ENTER key is
pressed to pass over a field, the cursor automatically moves to the next field.
Mixer/^T T ^ ^ ^ t0ge",er ^ «*»««• for the
cTJ : " ^ ^ ^ —by-screen corresponding
dosely to the pa^e-by-page order of the various Exposure Survey Forms. Items from the
posure Survey Forms that contain written text-lite descriptions, such as "Describe Application
ftocedure," are no. input into the database, m all cases, the daa entry operator must input only
the datum specified by the screen prompt,
2-3.2.1 Data Entry Prmrt
A typical data entry prompt is as follows:
01 Study Code [....]
2-7
-------
PHED Data Entry Menu
1.
Enter
Applicator Exposure Fora data
2.
Enter
Mixer/Loader Exposure Fora data
3.
Enter
Flagger Exposure Form data
4.
Enter
Combined Mixer/Loader/Applicator Exposure For* data
Press the nuaber of your selection
L
Return to PHED Main Menu
FIGURE 3. PHED Data Entry Menu
2-8
-------
The number "01" is the prompt number. Every prompt on the screen has a unique number.
The phrase "Study Code" is the field or datum description. The two brackets ("[" and "]")
define the data entry window. Data can only be entered between the two brackets, and the
cursor will only appear within the data entry window. The four periods between the brackets
are the data entry mask. Serving as a place holder for characters to be entered, the mask
indicates the maximum number of characters allowed for input
Each field has specific limitations (for example, numerics only, alphabetic characters
only, etc.). The mask for the field may also give information concerning the types of
inputs required. For example, a mask in the form "NN.N" means the input datum is expected
to be numeric, can contain one decimal place, and cannot be larger than 99.9. The rfatum is
typed in over the mask, and only those characters typed in from the keyboard are interpreted as
input by the computer.
There are two modes of operation for using a data entry screen: the "data entry mode"
is the normal input mode, and the "edit mode" is for changing previously entered records. The
user may employ the edit mode to edit records that he/she has added; however, there is a
data-protect program to prohibit editing of records which are part of the generic PHED. When
the user is in the data entry mode, the cursor will move from window to window sequentially
through the entire screen. Appendix A contains reference tables for PHED data entry screens.
The cursor will exit one window and move to the next window when one of three conditions
exist: (1) the entered data are in the correct format and fill the window exactly; (2) the entered
data are in the correct format and the ENTER key is pressed; or (3) the window is left empty
and the ENTER key is pressed. An exception to the third condition of leaving a null field is
that there are three fields which must not be left empty: the first three fields on the first data
entry screen (Study Code, Worker ID, and Work Cycle) are unique identifiers for that record
and must be entered on the first screen. The record number in these three fields will
automatically be carried over by PHED to each succeeding screen of the series for that record.
2-9
-------
2.3.2.2 HELP Screens
Individualized help messages are available for every prompt on every screen. Pressing
the Fl key will generate a help screen pertaining specifically to that field where the cursor is
displayed. Figure 4 is an example of a HELP screen; it describes the Study Code Number,
which is the first field on Screen 1. The help messages cover data entry parameters, siich as
whether the anticipated data must be alphabetic characters or numeric. If the data to be input
must be numeric, then there will be a range of acceptable values and a maximum number of
decimal places accepted by the screen. Revelation* will accept and store numbers up to
four decimal places for computations and data display.
2.3.2.3 Automatic Field Advance
If the user enters enough characters to completely fill the data entry madr the field value
is recorded, and the cursor automatically moves to the next field.
2.3.2.4 Multiple Choice Data
Several items on the Exposure Survey Forms are multiple choice items. One example
is the "Action of Pesticide" field (under the Compound Identification heading on the first page
of all the Exposure Survey Forms). The Action of Pesticide can only be one five types:
(1) Fungicide, (2) Herbicide, (3) Insecticide, (4) Fumigant, or (5) Plant Growth Regulator.
When entering this datum into the screen, the operator will input the number corresponding to
the Action of Pesticide. For example, if number 3, "Insecticide", is circled on the Exposure
Survey Form, the data entry operator will press the numeral "3" key when the cursor is in the
Data Entry Screen's Action of Pesticide window. The word "Insecticide" will pop up between
the curly brackets immediately to the right of the "Action of Pesticide" window, and the cursor
2-10
-------
STUDY CODE NUMBER
Enter the number found in the upper left corner of the Applicator
Exposure Survey Form. This number occurs in a box labeled "EPA Use Only."
The study code number may contain up to four digits. Entering four digits
will move the cursor to the next field automatically. If less than four
digits are entered, press the RETURN key to move to the next field.
End of Description. Press to return to prompt
FIGURE 4. Example Data Entry Hdp Screen
2-11
-------
will move to the next field, the "Liquid (lb ai/gal)" prompt window. The entry on the screen
for Action of Pesticide would then appear as follows:
07 Action of Pesticide [3] {Insecticide}
2.3.2.5 Multivalue Fields
, There are four instances where the multiple choice item on the Exposure Survey Form
may have more than one item circled. These include the Applicator's personal information field,
the "Type of Tank Additives"1 field, the "Type of Pesticide Diluents"1 field, and the
"Observed Type of Incidental Pesticide Contact" field
The incidental contact options include hand contact, observed face contact, and/or
observed contact with other body parts which occurred incidental to the pesticide handling and
application. The PHED Data Entry Screens have been programmed to handle multivalue fields
as a variation of the multiple choice field discussed in the previous section. For example, when
the cursor enters the window for field Number 11, "Type of Incidental Pesticide Contact", the
prompt appears as follows:
11 Incidental Contact [ ]
[ ]
[ ]
This multivalue prompt indicates the screen is ready to receive the first multivalue entry. If
there was no incidental contact, press the ENTER key; however, if there was incidental contact,
do not press the ENTER key until all types of contact have been entered. If hand contamination
(e.g., via a spill) has occurred for this worker replicate, then the numeral "1" key would be
pressed and entered on the screen, and the prompt would appear as follows:
lThese fields appear only in the Mixer/Loader and Mixer/Loader/Applicator files.
2-12
-------
11 Incidental Contact [01 >1] Hand Contact
The cursor moves to the next set of brackets, which indicates the screen is ready to receive the
second multivalue entry. If there is no second entry to be made, press ENTER and the cursor
will proceed to the next field or the "EDIT?" prompt, whichever is appropriate. However, if
face contact has also occurred for this worker replicate, then a numeral "2" would also be
entered on the screen, and the prompt would appear as follows:
11 Incidental Contact [02 >2 ] Face Contact
At this point, the screen is again ready to receive a third multivalue entry. The screen will now
appear as follows:
11 Incidental Contact [01 >1] Hand Contact
[02 > 2] Face Contact
[03 >.]
Thus, all three multivalue entries are displayed for Incidental Contact; however, for the other
three multivalue entry fields, not more than two entries will be displayed on the screen at any
time. The multivalue entry protocol is terminated at any time by pressing the ENTER key at
any of the numbered multivalue prompts.
2.3.2.6 Comments Field
A free-form field has been developed for the end of the last screen for each record. This
comment field will display up to 320 characters, including spaces and punctuation.
In the data-protected records, only the first four lines of the comment field are available
for printing via the Print Survey option or for viewing on-screen in the data entry "edit mode."
Although the comments field may be viewed, the data-protected records are not available for
2-13
-------
editing. In addition, these comments fields are not in a format which is available for subsetting
nor in a format which, in the Matrix Print or Browse procedure, may be displayed on-screen
or printed.
The comments field has been developed by invoking a Revelation®-provided text editor.
Thus, the field will scroll down automatically when entering information, and in reading
information in a non-protected record, the field may be scrolled down or up by n«ng the arrow
keys. The comments in the protected records available through PHED were written and entered
by Versar specifically for each record within each EPA Study Code Number. The comments
are intended as concise statements appropriate to that record concerning important descriptor
information not captured elsewhere on the Data Entry Screens.
2.3.2.7 "EDIT?" Prnmnt
The "EDIT?" prompt appears at the bottom of the screen. At this point, the user may
enter the number of a field to either add new data overlooked during data entry or to edit
existing data in the field which was erroneously entered. After editing that field, the cursor will
return to the "EDIT?" prompt at the bottom of the screen. (See Section 2.3.3 below.) During
the data entry mode, pressing the ENTER key at the "EDIT?" prompt will advance the user to
the next data entry screen.
2.3.3 Editing Saved Records
The user cannot edit data records distributed with the PHED package. Siting of the
Data Entry Screens is only for making changes to user-entered data (i.e., editing that data which
had been input, filed, and stored in PHED by the user during the data entry mode). At the
completion of the data entry_ mode for each screen, the user automatically drops into the grfjt
mode. The edit prompt will appear at the bottom of the screen:
"EDIT?"
2-14
-------
Pr0CeedS ^ lypin8 the data field number corresponding to the data he/she wishes
to edit and presses the ENTER key at the "EDrr>" nmm^ sue wishes
data entiy window for data field number selected after whi h th ^11101 ^10 ^
m,H. Mlected' ^ Wluch "* appropriate changes can be
If no changes are necessaiy at the end of data entry for a semen ». . _
at thP Nprtrr>« screen, press the ENTER key
"IT Prompt, and the next screen will be called. Pressing the ENTER key at thl
EDIT? prompt for lthe last screen in the dam entty mode will save that record along with any
ges were made. After that record is filed, the first screen in the screen sequence will
not^tr""^*^»^^^.^<^.«cordwill
In the edit mode for existing records, pressing the ENTER key at the "EDrP" prompt
results in an additional line appearing at the bottom of the screen as foUows:
"Press for the next screen, (S) for starting screen
or number of screen?" '
^-Proceed to thenext screen to continue editing this previously-entered dam record, press the
ENI^key. Pressing the S key followed by the ENTER key will take the user out of the edit
mode for that:record, will no, store any editing done on that re**, (the origin record win be
re-filed), and will return the user to the first screen for that data entry file to await data entry
for a new record. 3
To edit a previously-filed record, access the appropriate data file by calling up the data
entry screens in the usual fashion through the Data Entry Mdiu (Figure 3). the cursor will
appear in the Study Code window. Type in the Study Code, the Worker ID, and the Work
ycle numbers for the record"® be edited. Since this is an existing record, all the fields on the
screen after Work Cycle will automatically become filled with the data that were previously
2-15
-------
altered for this record, and the "EDIT?" prompt will appear at the bottom of the screen. For
example, to correct the tank capacity field on the second screen in the Mixer/Loader screen
sequence, the user will press the "ENTER" key at the "EDIT?" prompt By pressing ENTER
or the numeral "2" followed by ENTER, the second screen win appear. "Tank/Hoppers
Capacity" is Field Number 8 on screen 2, so the user would input "8" followed by ENTER, and
the cursor will move up to the data entry window for prompt number 8. The corrected data
would then be entered. When the "EDIT?" prompt reappears, if no other fields required editing,
the operator would input a series of two ENTER keys in order to exit the second screen and
then a series of ENTER keys to proceed to the end of the Mixer/Loader screen sequence and
save the changes.
2.4 Data Analysis
The structure of the Data Analysis Segment of the PHED System is shown in Figure 5.
This Segment of PHED is accessed at the Main Menu (Figure 2) by choosing the option "Data
Analysis." The "Data Analysis: File/Subset Selection" menu appears (Figure 6), and the user
then selects one of the four data files (Applicator=APPL.FILE; Mixer/Loader=MTXT n ftt p-
Flagger=FLAG.FILE; or Mixer/Loader/Applicator=MLAP.FILE) or an existing subset. To
select an entire file for the data analysis pathway, type the name of the file (as shown above and
as listed on the top half of the screen) and press the ENTER key. The same procedure is used
for selecting an existing subset; that is, type in the name of the subset and press ENTER. A
listing of the existing subsets b^ins at the bottom half of the screen in the subset directory; if
the directory extends beyond the half-screen, use the page up and page down keys to scroll
through the directory to view the complete listing.
After a data file or existing subset has been typed into the selection field, the Subset
Operations menu (Figure 7) appears at the bottom half of the Data Analysis Section: File/Subset
Selection screen. The user then selects one of the following: 1) create a new subset; 2) perform
exposure calculations; 3) perform statistical calculations; 4) select specific variably to be
2-16
-------
U
I
I
1*
i!
!¦
i
ii
i!
if
ii
11
i!
i
I
1.
i
I
ii
ii
S
3
O
s
CO
E
«
>»
(0
«
«
>»
"3
e
<
«
«
a
Q
m
in
S
3
2-17
-------
DATA ANALYSIS SECTION: File/Subset Selection
Please select a file or subset from the list below.
Naae:
MLAP.FILE APPL.FILE FLAG.FILE NIXLD.FILE
Use Hoae, End, Page Up and Page Down keys to viev subsets
<< Subset Directory >> Page 1 of 1
0447.APPL 0447.MLAP 0447.MLOD
0460.APPL 0460.MLOD 0468.APPL
AIR.APPL AIR.B.APPL AIR.OPEN.APPL
HEADAIR.APPL HEADCHEST.APPL HEADHANDS.APPL
NOT0448.MLOD OPEN.MIX.MLOD OPEN.MLOD
- Exit to PHED Main Menu
FIGURE 6. Data Analysis: File/Subset Selection Menu
2-18
-------
DATA ANALYSIS SECTION: Filt/Substt Scltction
Name: APPL.FILE
MLAP.FILE APPL.FILE FLAG.FILE MIXLD.FILE
Nana found, proceeding normally ...
« Subset Operations »
Please choose an operation to be performed on APPL.FILE
1. Subset
2. Exposure
3. Statistics
4. Browse/Print
5. Specifications
6. Delete
7. Print File/Sub
8. Print Comment
9. List/Count
- Return to File/Subset Selection
FIGURE 7. Subset Operations Menu
2-19
-------
browsed on the screen or printed on the printer; 5) view the specifications used to create the
current subset (not available for complete data files, only subsets); 6) delete the current subset
(not available for complete data files, only subsets); 7) print the data in all the records in the file
or subset in a format compatible with the Exposure Survey Forms (this option was developed
for obtaining hard-copy of entered data to permit quality control checking and verification of the
data entry process); 8) print comments; or 9) list/count file or subset record IDs. The user
selects one of these nine options by placing the cursor on the option with the up or down arrow
keys or by pressing the number of the option, then pressing the ENTER key.
2.4.1 Creating Subsets
The data analysis calculations may be performed on the entire data file or on a subset of
records from one of the files. The subsets are created by using the Subset Facility of PHED.
Because of the complexity of the subsetting procedure, it is suggested that the user should read
this section in its entirety prior to creating a subset. The screens and special keys are described
in general in the next two sections, the third section provides a step-by-step process for
variables and creating a subset.
2.4.1.1 Subsetting Screens
The first screen of the Subset Facility of PHED is called the "Subsetting Pathway"
(Figure 8). This screen provides a flow diagram for creating a subset. In general, a specific
variable is selected from one of the four pages listing variables; the user indicates whether to
include or exclude all records meeting the criteria placed on the variable in subsequent steps of
subset creation. The user will then specify relations and values for the variable. More than one
variable may be selected in the creation of a subset. Once the subset specification is complete,
the user provides a name if thcsubset is permanent, or a name is automatically given by PHED
if the subset is temporary. ~ ~
2-20
-------
<3> SUBSET FACILITY: Subsetting Pathway
Subsetting: The Applicator File
¦Variable Selecting
- Page 1
- Page 2
- Page 3
- Page 4
^^-Riteria^™ -> .in i"Hirer
1 ' SCLUDE I——I ~ Relation & I
J - EXCLUDE | j VALUE(S) I
<-
I Peraenant? —I 1 =>
I Temporary? —I I
I "UBSET—J ->
r-MORE CONDITIONS.
— YES AND
— YES OR
— NO DONE
<-
aae Subset.
» Help.
T
SUBSET FINISHED
J
¦ Specification.
• Accept.
FIGURE 8. Subsetting Pathway
2-21
-------
The Subsetting Pathway Screen is the main screen by which the user enters the Subset
Facility; and at each step in the subsetting procedure, PHED will return the user to this
Subsetting Pathway screen from the other screens of the Subset Facility. When the user initially
enters the Subsetting Pathway screen, the cursor is flashing on the Variable Selection box. To
define the first variable in the subset, the user selects one of the four pages (Figure 9). The
order of these variables corresponds to the order of the data elements on the Exposure Survey
Forms and on the Data Entry Screens.
Once the variable is selected, the next screen to appear is the Specification Screen
(Figure 10), where the user defines the restrictions to be placed on the selected variable. There
are three parts to this screen, "Relation," "Value," and "More." In the "Relation" part of the
screen (the left-hand column), the user chooses how the variable is to relate to the value(s). In
the "Value" part of the Specification Screen, the user specifies the value (or values) for the
variable which the records in the data file are required to meet for inclusion (or exclusion) in
the subset. In the "More" part of the screen, a variable may have one or more relations defined
on it. If additional relations are to be defined on a variable, this "More" facility provides a way
to select another relation and value on that variable. Figure 10 shows an example in which two
values have already been specified and PHED is awaiting a third value or giving the user the
option of concluding the relation-and-value specification for that variable. Figure 11 is an
example of a Specification Screen for the variables included in PHED as values.
In these variable selection screens, there are two kinds of Help Screens, those that
provide guidance at a given step in the Subsetting Pathway and those that assist the user in
Variable Selection. Both types of Help Screens are viewed by pressing the F1 key. The content
of the Help Screen will depend on where the user is in the subset specification when hrfp is
requested. When the user is in any of the boxes on the Subsetting Pathway Screen, one of the
series of Help Screens concerning that step in the subset specification procedure is given; thus,
there is a different Help Screen for each box on the Pathway Screen. A specific Help Screen
for a given variable in the variable Selection Screens can be invoked by highlighting the variable
2-22
-------
<3> SUBSET FACILITY: Variable Selection - Applicator File
Help.
¦ Specification-
¦ Accept.
FIGURE 9: Subset Variable Selection
2-23
-------
<3> SUBSET FACILITY: Variable Selection - Applicator File
Application Method
Rate (lb ai/acre)
Rate (9al/acre)
Total lb ai Applied
Total I Acres Treated
Final Nix (lb ai/gal)
Total # gal Sprayed
Vehicle Make 6 Model
i Tank Applications
Cab Type
Speed
Page 2 of 4
AI - cont.
Applicator Make/Model
Disc t
I Nozzles
Nozzle Type
Nozzle Model/No
- Nozzle Pressure
Boob Height
Svath Width
Boon Length
Shank/Prob Type
Depth of Inj/Incorp
»< Fl> Help.
¦ Specification-
¦ Accept-
FIGURE 9: Subset Variable Selection (continued)
2-24
-------
<3> SUBSET FACILITY: Vsrisble Selection - Applicator File
Weather Data
High Humidity
Low Humidity
High Temperature
Lov Temperature
Wind Speed
************
************
Wind Direction
Personal Info
Study Participant
Years Experience
Page 3 of 4
Airborne Exp Data
Technique
Air Sampling Time
Air Quant. Limit
Amount Collected
Volume Collected
Initial Flow Rate
Final Flov Rate
Average Flov Rate
»< Fl> Help.
. Specification.
¦ Accept*
FIGURE 9: Subset Variable Selection (continued)
2-25
-------
<3> SUBSET FACILITY: Variable Selection - Applicator File
Dermal Exposure Data
Location Inside Outside
Right Left LOC Right Left
Head
Neck Front
Neck Back
Both Hands
Hands
Upper Arm
Forearm
Shoulder
Chest
Back
Hips
Thigh
Shin
Calf
Ankle
Foot
I
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
0
0
0
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
0
0
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
0
0
0
0
+
+
+
+
+
+
+
+
+
+
+
+
Page 4 of 4
Sampling Time
Measure Method
Hand Measuring Method
Avg Dermal Quan. Limit
Avg Hand Quan. Limit
- Incidental Contact
¦ Help.
• Specification.
¦ Accept.
FIGURE 9: Subset Variable Selection (Continued)
2-26
-------
Choose Variable Relation to Value
Choose Value(s) to liait Variable
Study Code
STUDY.CODE
Equal to
Not Equal to
Less than
Less than or Equal to
Greater than
Press for a description
of the acceptable value range
and units of this variable.
Greater than or Equal to
Starting with
Not Starting with
Ending in
Not Ending in
Containing
Not Containing
VALUE 1 > 0430
VALUE 2 > 0435
VALUE 3 >
Help Specification—__ Accept
FIGURE 10: Subset Variable Specification
2-27
-------
Choose Variable Relation to Value
Choose Value(s) to limit Variable
Pesticide Action
ACT.PEST.CODE
Equal to
Fungicide
Not Equal to
Herbicide
Less than
Insecticide
Less than or Equal to
Fuaigant
Greater than
Plant Growth Reg.
Greater than or Equal to
Starting with
V
Not Starting vith
Ending in
Not Ending in
Containing
Not Containing
Help SpecificatioDaMMMMi^B Accept
FIGURE 11: F.xamplr Coded Variable Selection
2-28
-------
screen is accessed by pressing in E2 key.
2.4.1.2
Special lfgy
The escape key, , may be pressed a any point in the subsetting procedure to
key provides a n^od of returning
Va^°sZr,!COrreCt,niaakeS- ^ ^ ^ - be used to return back to the
variable Selection box which i« th* »*,*_, __ ...
TonHn ¦ , u- v position on the Subsetting Pathway Screen
Continuing to hit the key will return the user to previous screens, such as the data
subset menu and the PHED main menu. After moving back to Uus Variable Selection Bo^.the
user may again press or press to exit the Subset Facility without creating a
spec^fi °ti r ^ h*"1 VailableS Sele°ted' ^ ^ Press to bypass the subset
specification loop and pass directly to the Permanent-Temporary Subset Box.
(B)
key is the Help key and is active a. aU times during subset specification.
Th e Help Screen which appear, depends on where the user is in the subset specification when
l-T* DUrin8:Variable «"*•. ^e Help Screens for specific variables are
slightly edited versions of the Help Screens found in the Data Entry Segment of PHED. At all
2-29
-------
<3> SUBSET FACILITY: Current Subsetting Specification Page 1 of 1
CMD Pos Variable Selection Criteria
1 Study Code With Study Code Equal to 0430 and
2 Pesticide Actio With Pesticide Action Equal to 1 2 and
3 Study Location With Study Location Equal to "CANADA"
Press to return.
FIGURE 12: Example Subsetting Specification Screen
*
*
2-30
-------
other times in the Subsetting procedure, pressing provides a Help Screen specific to that
part of the Subsetting Pathway.
(C)
Pressing the key provides a listing of the current Subset Specification. The
listing is comprised of the position of the variable in the specification string, the internal name
of the variable in the PHED dictionary, and the specifications) defined for the variable. If
is pressed when the cursor is in the Variable Selection List, 'Variable Selection' box,
or when the subset specification is complete, the Delete option CD" key on the keyboard) is
activated. At this point, one or more variables may be removed from the specification by
moving the cursor in front of the desired variable using the up and down arrow keys and
pressing the "D" key on the keyboard. The delete operation is then executed by pressing the
ENTER key. The "D" key and Delete operation is removed by again pressing the DELETE
key. This will clear the "D" and Delete operation; and the specified variable will not be deleted.
The Specification Screen is exited by pressing the ESCAPE key.
i
2.4.1.3 Steps in Sybset Specification
The following discussion pertains to each box found on the Subsetting Pathway Screen
(Figure 8).
2.4.1.3.1 Variable Selection Box
The variable selection box is the starting point for the list of variables for subset
specification. The variables are displayed (Figure 9) and arranged in the order in which they
are found on the Exposure Survey Forms and the Data Entry Screens. The cursor is moved up
or down on the menu to specify the page of choice, then pressing the ENTER key brings up the
variable list for that page. At this point, any page can be entered using the PAGE UP or PAGE
DOWN key.
2-31
-------
A variable is selected by moving the cursor with the up or down arrow to the desired
variable so that the variable is highlighted, then pressing ENTER. As the cursor moves over
the variables, the name of the variable is displayed in the screen's upper left corner. If the
variable has already been selected and is included in the current subset specification, the word
SELECTED' is highlighted next to the data name at the upper left of the screen.
If a previously selected variable was to be chosen, a prompt would appear at the bottom
of the screen asking if the user wants to modify the specification for that variable. If the answer
is Yes , the previous specification is deleted, and a new application for the variable must be
selected. If the answer is "No", then the user moves on to select a different variable.
Pressing will display the Help Screen for the variable in which the cursor is
currently positioned. Pressing < F2 > provides a listing of currently selected variables and their
specifications. At this point the user may delete a specified variable by the use of the "D" key.
The key exits the Variable Selection List without specifying any variables.
2.4.1.3.2 Criteria Box
Selecting the 'INCLUDE' option means that all records of the data file meeting the
criteria applied to the variable will be included in the subset. The 'EXCLUDE' option means
that all records of the data file that meet the criteria applied to the variable will be
from the subset. PHED has been programmed for 'INCLUDE' to be the default option.
2.4.1.3.3 Select Box
The Select Relation and Value(s) Box is the starting point to define the specification(s)
to be applied to each variable selected.
To define the specification, a relation must first be selected. This is done by moving the
cursor up or down to the desired relation on the menu, then pressing the ENTER key. The next
2-32
-------
step is to enter the value or values which are to be associated with this relation. For some
variables, the set of possible values will be displayed, such as "coded" variables where a number
was used in the Data Entry Screens to represent the description. For example, Pesticide Action
is a coded variable, and the possible specification values (Figure 11) are displayed so the type(s)
of pesticide may be chosen.
A specific value relationship from the left-hand column is selected by pia™£ the cursor
over the desired name and pressing the ENTER key. If the relation 'EQUAL TO' or 'NOT
EQUAL TO' was selected, more than one value may be selected. For coded variables,
two relations are the only reasonable ones to use. Thus, phki* has been programmed not to
allow the user to select 'STARTING WITH", 'ENDING IN', 'CONTAINING' and their
negatives for coded variables. When more than one 'EQUAL TO' value relationship is selected,
an 'OR' condition exists between the two values. This section of the Subset Pathway is exited
by pressing the ENTER key twice in succession.
When the variable selected is not a "coded" variable, thai a value must be altered at the
"VALUE 1>" prompt. Pressing the Help key () provides information about the
variable, its range of possible values, and its limits.
When 'EQUAL TO', 'CONTAINING', 'STARTING WITH', 'ENDING IN' or their
negatives have been selected as relations, then more than one value may be given. The relations
'CONTAINING', 'STARTING WITH', and 'ENDING IN' may be used with alphanumeric
values only.
At the end of each value specification, a prompt is produced asking the user if another
relation should be defined on the current variable. If so, the user should move the cursor to the
'AND' option if the subset selection is to further restrict the new relation, or to the 'OR' option
if the new relation is to expamtthe subset selection. The user chooses the 'DONE' option when
subset specification for the current variable is finish^,
2-33
-------
The following two examples illustrate these procedures:
Example 1. To create a subset where the study code is 0420 £ Study Code s 0435 and
where the head patch values are greater than 0.
Variafrte Relation Value More?
Study.Code Less Than or Equal To 0435 AND
Greater Than or Equal To 0420 DONE
HEAD Greater Than 0 DONE
Example 2. A subset where all records have a Dermal Qiiantifiratirm l imit and the
were performed in ranaHa
Variable Relation Value More?
DERM. LIMIT Not Equal To Null* DONE
SITE.LOC Containing Canada DONE
*NULL is a special value used to specify variables with no value asgignpH
2.4.1.3.4 More Conditions Box
After the subsetting criteria have been defined for a variable, the user proceeds with the
More Conditions Box. At this point requesting 'YES AND' or 'YES OR' allows the selection
of another variable for further subset specification. Choose the 'YES AND' option to further
restrict membership in the subset with criteria to be specified for the new variable. Choose the
'YES OR' option when you want to expand the subset selection with the new variable.
When choosing the 'NO DONE' option in the More Conditions Box, the specification
cycle is terminated, and the user proceeds to name the subset.
2-34
-------
2.4.1.3.5 Permanent or Temporary Subset and Name Boxes
The last three boxes on the Subsetting Pathway screen concern the naming the subset.
The subset may be permanent or temporary. A permanent subset is given a name by the user
which must be 25 characters or less in length. Names may contain numbers, letters, and
charactere- Other characters will be ignored, and spaces can not be used. If the user
chooses to use a subset name which has already been assigned to another subset, an appropriate
message is shown; and the user has the choice of assigning a different name to the new subset
or using the existing name to write over the old subset However, the preferred method of
writing over or deleting an existing subset is to select the "Delete" option on the Subset
Operations menu.
Temporary subsets are given the name "TEMP.NAME" by default by the PHED system.
If this happens to be the name of an existing subset, the existing subset will be replaced by the
new subset, so this name "TEMP.NAME" should not be assigned by the user to a permanent
subset which the user wants to save.
Each subset created, whether permanent or temporary, automatically receives an extension
appended by PHED to its name identifying the data file from which it originated. There are up
to four temporary subset names which may be assigned by PHED, one for each data file. The
following extensions are used:
".MLAP" - Mixer/Loader-Applicator File;
".APPL" - Applicator File;
".FLAG" - Flagger File; or
".MLOD" - Mixer/Loader File.
The Delete" option on the Subset Operations menu provides a way to rfrfety subsets
which are no longer needed.
2-35
-------
2.4.1.3.6
Subset Finished Box
When the user has moved the cursor through the Subsetting Pathway Screen to the Subset
Finished Box, press ENTER to begin the process of creating the specified subset. The user,
at this point, may press < ESC > to move back through the Subsetting Pathway and to
changes in the specification; or the user may press to review or delete variables in the
specification. If no problems are encountered in subset formation, the subset is saved, and the
user is returned to the Subset Operations menu. The user will then be given the choice of using
either an entire data file; the original, parent subset; or the newly created subset ("O" for old,
"N" for new). PHED will not permit the user to continue until this question is answered.
There are three potential problems which may result during the subset formation
procedure:
1. No records in subset;
2. A subsetting specification is longer than 255 characters; or
3. An invalid subsetting specification.
If any of these conditions arise, no subset will be created or saved.
If condition (1) occurs and there are no records in the subset which has been specified
by the user, the screen will indicate "... zero records. Subset not saved." At this point, the user
may press any key to return to the Permanent-or-Temporary Subset Box. By then pressing the
key, the user may move back through the Subsetting Pathway Screen; or by pressing
, the user moves directly to the Subsetting Specification Screen, where the user may
delete those variables which had restricted the subset to the null set containing zero records.
If condition (2) occurs _and there are more than 255 characters in the subset specification
string (the maximum length accepted by the Revelation* software), then the screen will indicate
that the string exceeds 255 characters and will specify at which variable in the string that the
2-36
-------
overrun occurred. By pressing any key, the user returns to the Permanent-or-Temporary Subset
Box and may use the key to modify the last variable specified or may use the
key to delete as many variables as desired to bring the subset specification string back within
the 255 character limit.
Condition (3), invalid specification procedures, will occur if the subset procedures have
been invalid and Revelation* is unable to proceed with subsetting. In this case, the user will be
informed by the screen concerning which variable or relation caused the subsetting procedures
to terminate.
2.4.2 Exposure Analysis Pathway
After selecting a specific Data File, creating a subset of records from that Data File, or
retrieving an existing subset, the user may then select the Exposure Path from the Subset
Operations menu. The user next sees the "Algorithm Selection Menu" (Figure 13). The user
then selects one of the three methods of expressing exposure Hata- inhalation (by pressing key
"1"); total dermal exposure (by pressing key "2"); the combined value of inhalation and total
dermal exposures (by pressing key "3"); or return to the Subset Operations Menu (by pressing
F9 or ESCAPE). For this Algorithm Selection Menu and most menus in the Exposure Path that
require a user to enter the desired menu number, enter the desired numeral but then de not press
the ENTER key afterwards. PHED will respond to one input character at a time; extra input
characters will result in beeping at the next menu in sequence, or may cause that next menu in
the sequence to flash by rapidly without allowing the user a chance to read the screen or respond
to any menu queries.
For any menu, the user may return back through the previous menu chain by hitting
, the escape key, or the key. The only exception to this backwards movement
is when one of the algorithms is functioning and the "Percent Completed Bar Chart" is on the
screen with the barometer-like bar moving across the screen from left to right showing the
percentage of data records processed. Do not press any keys while this operation is functioning.
2-37
-------
CALCULATED EXPOSURE ALGORITHMS - APPLICATOR DATA
Select an exposure algorithm:
1. Inhalation exposure
2. Total dermal exposure
3. Combined inhalation and dermal exposures
Press the number of your selection
Return to subset operations menu
FIGURE 13. Exposure Algorithm Selection Menu
2-38
-------
He user may** escape 5^ ^ ^
htte'«sai"^¦»¦»-»- *» me ^ Ca1cUfctir^ tzr***
»«. * asked to
-—:^^ r^r3) t
niate a selection, press the numeric key cotresponding to the desdnd^hcrice ^OT^hTl^x /
I^der/Appiicator file, the user has three options for ^ lbs ^ ^ *» * ^
iZT^r:m^otw^- 2»
an rt„ _ records axe being processed and, then will see
an °n-scree„ Data Report "Summary Statist for mhalation Enures- (Figure M)£
reen shows the normalized inhalation algorithm results, expressed in nanograms Tie data are
Zr ofT "T ^ ^ — of variation, geom^n^ Z
J^T ^SerVaa0nS °° C°bs--) * *• subset having sufficient data for those
in the bT '°ICa^e "" mhalati0n CIp0Sure among the total number of records
«-J12- fc 95 ^ «—• «« — ~• - -
t ^ mhak°0n ^ ™ algorithm computations is shown on the screen
— •oo^eforittres[«tivefieM„nti« screen, the field is filled with as***
MM bottom of the screen, ti* ^ of the subset, the name of the da* file, JT£
umber of records contained in the subset are displayed. There is also a menu bar at the bottom
b- of the screen . allow the user to convert * M ^ ^
2-39
-------
SUMMARY STATISTICS FOR INHALATION EXPOSURES
DISTRIB. NANOGRAMS PER LB AI SPRAYED
TYPE Median Mean Coef of Var Geo. Mean Obs.
EXPOSURE Lognormal 4000 9153.6389 209.96 4044.5376 40
95% C.I. on Geo. Mean: [402.3046, 40661.4436]
Number of Records: 40
Data Pile: APPLICATOR Subset Name: AIRBLAST.OPENCAB.APPL
**~ Press 'P' to print this report ***
Convert LB AI to KG AI EXIT
FIGURE 14. Example Inhalation Exposure Summary
Statistics Screen
2-40
-------
T°T screen' "^y either choose the EXIToption on the menu bar
at the bottom of the screen or press the key or key.
Tbt inhalation algorithm is a subroutine which processes the data individually ftom each
subset record selected. Further details concerning the inhalation algorithms are descnbed i.
Section 3. If respirator pads or patches are the reporting mettod for a record's "Amount of
emical Collected G«g)-, then this microgram value is normalized. If the normalized variable
is less than 0.0001, then the variable is assigned the value of 0.0001. (This is the smallest
non-zero number allowed by Revelation*.)
If a personal air pump or a fixed location air sampler was used to collect the data for the
record's "Amount of Chemical Collected <«)•, this mass of pesticide is multiplied by the human
ventilation rate divided by the average flow ra* of the pump
-------
appear to be-an acceptable method for statistically weighting these unequal for the
different body parts. Instead, each record is processed as a separate replicate; this method is
more acceptable statistically than treating as separate groups each record from the different field
studies. (For example, it would not be appropriate to perform separate statistical analyses for
each different Study Code Number in a given data file, and then to composite the information
for each record in a study based on separate body parts to determine the total exposure for each
of the individual Study Code numbers.)
To select the Total Dermal Exposure algorithms, the user presses the "2" key on the
"Algorithm Selection Menu" (Figure 13). At this time, the user will be asked to select a method
of normalization. The available choices are: 1) time; 2) total lbs AI; 3) total lbs Al/time;
4) rate lbs Al/acre/time (not availabe for the Mixer Loader file); and 5) none. To a
selection, press the numeric key corresponding to the desired choice. For the Mixer/
Loader/Applicator file the user has three options for total lbs AI: 1) total lbs AI sprayed;
2) total lbs AI mixed; or 3) average of sprayed and mixed. After selecting the normalization,
the "Dermal Exposure Scenario Specification menu" appears on the screen, offering the user the
choices shown in Figure 15. The scenarios include: one "No clothing" algorithm; four
"Normal Work clothing" algorithms (Figure 16); and six "Protective Clothing" algorithms
(Figure 17).
PHED has been programmed to analyze separately the data for each body part to
determine whether the body part data are normally distributed, lognormally distributed, or
whether the distribution pattern is random (other). That is, the data are analyzed to determine
the frequency distribution for the arithmetric (untransformed) data and for the geometric
(log-transformed) data. These frequency distribution computations are utilized to determine the
most appropriate ("best-fit") procedure for processing each body part in the total dermal
exposure calculations. (Further details of these frequency distribution analyses are described in
Section 3,) - - *
2-42
-------
CALCULATED EXPOSURE ALGORITHMS - APPLICATOR DATA
Select a denial scenario:
1. No clothing (total deposition)
2. Normal work clothing
3. Protective clothing
Press the nuaber of your selection
Return to algorithm selection aenu
FIGURE IS. Dermal Exposure Scenario Specification Menu
2-43
-------
CALCULATED EXPOSURE ALGORITHMS - APPLICATOR DATA
Select a nornal work clothing scenario:
1. Long pants, short sleeves, no gloves
2. Long pants, short sleeves, gloves
3. Long pants, long sleeves, no gloves
4. Long pants, long sleeves, gloves
¦
Return to denial scenario selection menu
FIGURE 16. Normal Work nothing Menu
•2-44
-------
CSLCOLATED EXFOSME ALGORITHMS - APPLICATOR DATA
Select , protective clothiog seen"™?
1.
2.
3.
4.
5.
6.
Protective
Protective
Protective
Protective
Protective
Protective
overall
overall
overall
overall
overall
overall
over long pants, short sleeves, no gloves
over long pants, short sleeves, gloves
J0ng pants' lon9 sleeves, no gloves
long pants, long sleeves, gloves
over no clothing, no gloves
over no clothing, gloves
Press the number of your selection
Return to dermal scenario selection
aenu
FIGURE 17. Protective fTinthing
2-45
-------
As previously discussed, the patch data far total dermal exposures-is processed differently
from the inhalation algorithm (i.e., data are processed by computing . separa* algorithm far
eac record). Instead, the patch data from each individual reoord in the subset are read by the
Total Dermal Exposure Algorithm record-by-record to select the data from the fields of interest
For ead, record for each patch for each body tocation, the data from the record are read
into drfferem arrays. If the patch data were entered during the da* entry stage and report in
e data ffles as "ND," the subroutine then extracts and stores a value equal to one-half the limit
of quantification entered for that record.
If patch data are entered in the data files as -C- in one field, the exposure data listed for
the oppos«e side of the body part are the -combined- patthda*. to is, two (normally left and
ngh ) patches were collected in the field, but were combined in the laboratory and only one
combined value was reported. In this case the subroutine will stare a value of on^-halfthepafch
v ue (i.e., average for that body part). "M" stands for middle patch location, and only the one
patch was analyzed to represent the average exposure to a body part. "M" values should occur
only for the chest and back body parts.
If patch data are missing, this body part location is processed as a null set because the
body par, was no. sampled in the field study, and this part of the record, data set is no, used
or exposure algonthm computations. (See Section 3 for further details of letter codes for the
body patch fields.)
All body part patch data which are non-zero (actual values in „gW or "ND- values
greeted for 50 percent of the limit of quantification values in w-W) are then normalized
If a vanable is less than 0.0001, a value of 0.0001 is assigned; tins is the small*. „0n-zero
number allowed by Revelation- Tie da* in these arrays are then processed to statistically
analyze the median, mean, standard deviation, variance, coefficient of variation, geometric
mean, and geometric variance for each body part Hese composite data for each body part are
then multiplied by the standard surface area of that body part. (See Table 1.)
2-46
-------
Table 1. Surface Azeas for Regions of the Adult Body
and Locations of Dermal Exposure
Pads that Represent These Regies
Region of
the body
Surface area (cm2)
of region
location of pad(s)
representing region
Head
Face
Back of neck
Front of neck®
Chest/Stomach
Back
Upper arms
Forearms
TTanHt
Thighs
Lower legs
Feet
1300*
650
110
150
3550
3550
2910
1210
820
3820
2380
1310
Shoulder, back, chestb
Back
Chest
Chest
Back
Shoulder and forearm/
upper arm
Forearm
Thigh
Shin
* Surface area for the head includes the 650 cm2 face surface area.
Exposure to die head may be estimated by using die mean of die shoulder, back, and riwtf
patches, or by using a head patch.
c Includes "V" of die chest.
Source: USEPA (1987) Pesticide Assessment Guidelines. Subdivision U. Applicator
Exposure Monitoring.
2-47
-------
For some body parts, surrogate data are used to estimate exposure for those records in
which the specific body parts were not monitored. For the front of the neck, the patrh^
for that record (individual worker) are an acceptable substitute; for the back of the neck, the
back patches for that record are acceptable; for the upper arms, all available upper arm and
shoulder patches for that record are averaged to obtain upper arm values; for the upper legs, the
patches from the hip and thigh are averaged for each record; for the lower legs, the shin, calf,
and ankle patches are averaged for each record. The user will not be aware that the PHED
Total Dermal Exposure Algorithm has averaged the available data for each of the individual
records in this manner, because the user is not notified how these particular are
processed during algorithm execution. (See Section 3 for further discussion of surrogate
data).
The head can be an important exposure site and the location of the patch representing the
head area may cause the exposure estimate to differ greatly. Thus, at the patch
Specification Menu (Figure 18), the user may elect to not use surrogate head exposure data.
The user has one of three options: 1) processing only the data from those records which have
actual head patch data; 2) processing only those records for which head patch data are missing
but for which chest, shoulder, and/or back patch data are present and can be used to estimate
the head exposure data; or 3) combining the existing head patch data with the mining head
patch/estimated head values.
Once the user has selected one of the three head patch scenarios, then the
percent-completed bar chart for calculation appears, and these calculations are completed. The
larger the number of records, the greater the time that PHED requires to process the algorithm
and report the results. The user should not press any keys while the dermal exposure algorithms
are being computed.
The Summary Exposure Report screen will appear after all exposure and statistics
calculations have been completed (Figure 19). There is a mode on the menu bar at the bottom
2-48
-------
CALCULATED EXPOSURE ALGORITHMS - APPLICATOR DATA
Select a head patch scenario:
1. Use actual head patches only
2. Use estimated head patches only
3. Use actual and estimated head patches
Press the number of your selection
Return to dermal scenario selection menu
FIGURE 18. Head Patch Specification Menu
2-49
-------
SUMMARY STATISTICS FOR
SCENARIO: Long pants, short sleeves,
PATCH
LOCATION
HEAD (ALL)
NECK.FRONT
NECK.BACK
UPPER ARMS
CHEST
BACK
FOREARMS
THIGHS
LONER LEGS
FEET
HANDS
TOTAL DERM:
DISTRIB.
CALCULATED DERMAL EXPOSURES
gloves
MICROGRAMS PER LB AI SPRAYED
TYPE
Median
Mean
Coef of Var
Geo. Mean
Obs.
Other
10.79
4267.7528
364.9147
22.7908
272
Other
1.305
445.3571
419.2135
2.2442
265
Other
1.111
195.0228
357.0261
1.5751
261
Other
90.792
1893.7108
211.7499
108.8949
72
other
17.395
2670.7656
516.3471
22.4363
127
Lognormal
51.83
1292.0954
282.4709
28.8862
112
Other
31.218
14515.3636
393.141
46.1998
252
Other
114.6
2413.3215
400.1578
60.8008
89
Other
74.256
1979.7321
338.6348
49.1954
89
Other
10.6667
364.6558
359.6542
15.7908
0
73
381.0199
403.9637
30037.7775
358.8143
Mean: Dermal:
[-227353.4613, 287429.0163]
Data File: APPLICATOR
ADD INHALATION
Number of Records: 282
Subset Name: APPL.FILE
CHANGE HEAD LB AI TO KG AI
*** Press ' P* to print this report ***
EXIT
FIGURE 19. Example Dermal Exposure Summary Statistics Screen
2-50
-------
of the screen to convert exposure estimates in "mg/lb ai" to "mg/kg ai," and vice versa,
applicable only to this method of normalization.
The scenario used to calculate these exposures is displayed as the second line, just below
the title. The data are displayed on each of these Data Report screens to show summary
statistics values for each body part location used in the algorithm, as well as the composite
calculated values for total dermal exposure in micrograms. The values reported for each body
part include the type of frequency distribution, the median, mean, coefficient of variation,
geometric mean, and number of valid observations (obs.) among the total nnmhrr of records in
the subset At the bottom of these columns of data for each body part, there axe combined
dermal exposure estimates for 1) the " most-appropriate method of expressing body part data"
(i.e., arithmetic means are used for those body parts which are normally distributed, geometric
mean for body parts which are lognormally distributed, and m«Hian« an ^ fOT those body
parts which are neither normally nor lognormally distributed); 2) total dermal exposure based
on only arithmetric means for each body part; and 3) total dermal exposure based on only
geometric means for each body part. (See Section 3 for detailed rfireii«inns 0f these
computations.)
In addition, the 95 percent Confidence Interval is reported for the.total dermal exposure
based on arithmetnc means. These Confidence Interval values are computed on a
statistical weighting procedure which incorporates the variability values for each irwati^
to obtain the composite variability of the total dermal exposure. (See Section 3 for a
discussion of the computation procedure for these Confidence Intervals.) Note: The 95 percent
Confidence Interval reported for the composite dermal exposure is inappropriate and shnuM be
ignored. In addition, die 95 percent Confidence Interval on the geometric mwan inhaiatinn
exposure is also reported on the combined Exposure Data Screen.
The data reporting for the Revelation* software package is to a maximum precision of
four decimal places (e.g., 104 fig or 0.0001 ng). Thus, in some cases a zero value may be
reported on these Data Report Screens, especially for the median, because the computed
2-51
-------
exposure value for that body part is less than the limit value for data reporting. Also, due to
maximum precision of Revelation* computation procedures only retaining four decimal places,
a small percentage error may be incorporated into the calculations. Conversely, if a data item
is too large for its respective field on the screen, the field is filled with asterisks, although the
value is stored properly. \
The menu bar at the bottom of the Total Dermal Exposure Screen may be used to add
the Inhalation Exposure and, thus, to generate a combined Dermal/Inhalation Exposure Data
Rqxwt Screen. The menu bar selection mode is moved by u-«ing the right or left anew key
keys, and the selection is activated by pressing the ENTER key. The inhalation rate used in
inhalation calculations will also be displayed. When die menu bar at the bottom of the screen
is placed on the "Add Inhalation" mode and the Combined Dermal and Tnhaiati™ Exposure is
already on the screen, the terminal will beep, and the word "Inhalation" will flash next to the
appropriate data.
Another option on the menu bar at the bottom of the screens is "Change Head," meaning
to change the type of head exposure option (Figure 19). The user will be able to use the menu
bar in this mode to change screens and view head exposure values and total dermal exposures,
with the "Head" label appearing as "Head (Real)", "Head (Est)", or "Head (All)" for the
exposures with only actual head patches alone, with only estimated head exposure when patches
are not present, and the combination of actual and estimated head data, respectively.
When the menu bar is on the "lb AI to kg AI" option, data conversions are performed,
and the corrected data appear on the screen. When the menu bar is in the "Exit" mode, this
option will allow the user to return to the Exposure Path initial menu, "Algorithm Selection
Menu. In each case, the menu bar mode selection option is executed by pressing the ENTER
key. Another method of returning to the "Algorithm Selection Menu" is to press the
key or the key to execute the escape option. While the exposure screen is being
displayed, pressing "P" will cause the report to be printed.
2-52
-------
Above the menu bar at the bottom of the Data Reports Screen for Total Dermal Exposure
and Combined Exposure, there are three entries. "Data File" indicates the data file from which
the data has been extracted (Applicator, Mixer/Loader, Mixer/Loader-Applicator, or Flagger).
Number of Records" indicates the total number of records contained in the entire subset, not
just those records having observations ("obs.") for specific body parts. "Subset Name" refers
to the subset on which the dermal and inhalation exposure computations were performed.
2.4.3 Statistical Analysis Pathway
Once the user has selected the statistics path on the Subset Operation Screen (Figure 7),
the user proceeds to a series of screens, starting with the "Statistics: Variable List Entry
Screen" (Figure 20). This screen gives instructions for the selection of variables and for the
statistical analyses to be performed on each variable. Pressing the "1" key will take the user
into the Variable Selection Lists Screens. Pressing the "2" key will direct PHED to raimw
the total body statistics using only data sets that are deemed to be "complete" according to
criteria established by the user. Pressing the "3" key will clear all currently selected variables
and take the user into the Variable Selection Lists Screens. Pressing the "4" key will return the
users to the Subset Operations Menu.
When the user enters the Variable Selection List screens (pages 2 through 4 of
Figure 20), the user may move the cursor within a page from one variable to the next or back
to the previous variable by using the "right arrow", "left arrow", "up arrow", or "down arrow"
keys. The user may move through the series of three "Variable Selection List" screens by using
the PAGE UP or PAGE DOWN keys.
Before selecting the variables on which statistics will be performed, the user must select
a normalization variable. Press the numeric key that corresponds to the desired choice:
2-53
-------
STATISTICS: Variable Selection List
Mixer/Loader Pile
_ Page 1 of
Cursor Movement and Variable Selection
Page Up - Move* cursor to the first variable on the previous page.
Page Down - Moves cursor to the first variable on the next page.
Right Arrow - Moves cursor to the next numbered variable.
Left Arrow - Moves cursor to the previous numbered variable.
Enter Key - Selects a variable and allows input of specification.
- C: Correlation; R:Regression; U:Univariate; X:ASCII File
~ PrM* Enter again to signify specifications are complete.
Delete Key — Deselects a selected variable.
Please choose one of the following options
1. Select variables for body part statistical analysis.
2. Recalculate total dermal statistics using "complete" data sets onlv
3. Clear selected variables.
4. Return to the Data Analysis entry screen.
Please Enter your choice:
Note: Units for statistics - Air amount of chemical. Hands fjg
Dermal body parts /jg/cm2
Body parts for "complete" data sets ug
Only DEPENDENT variables will be normalized.
— Variable Selection Completed -
-------
STATISTICS: Variable Selection List
Applicator File
Maximum variables/pathway : Corr.
1. Crop Height
3. Rate Lb AI/Acre
5. Total Lbs_AI Applied
7. Total f_of Acres
9. No. of Appl.
11. Nozzle Pressure
13. Svath Width
15. ***************
17. ***************
19. Low Huaidity
21. Low Teap.
23. Air Saaple Tine
25. Air Aat of_Chea
27. Ave Flov Rate
Page 2 of
6; Reg., 24; Univ., 6
2. Row Spacing
4. Rate Gal/Acre
6. Final Nix Cone
8. Total Gal Sprayed
10. Ground Speed
12. Booa Height
14. Booa Length
15. ***************
18. High Huaidity
20. High Teap.
22. Wind Speed Code
24. Air Quant. Limit
26. Air Volume
28. Dernal Saaple Time
Normalized by Total Lbs AI
For regression all variables on this page are independent except 125.
NOTE: '*• denote variables that are not members of this file.
Variable Selection Completed - to VAR LIST ENTRY SCREEN
FIGURE 20. Statistics Variable List Entry Screen (continued)
2-55
-------
STATISTICS: Variable Selection List
Applicator File
Maxiaun variables/pathway
Page 3 of 4
Corr., 6; Reg., 24; Univ., 6
Patch Inside Clothing
29. Head Inside
31. Neck Front Inside
33. Neck Back Inside
35. Upper Ara Inside
37, Shoulder Inside
39. Chest Inside
41. Back Inside
Foreara Inside
Hip Inside
47. Thigh Inside
49. Shin Inside
51. Calf Inside
53. Ankle Inside
55. Foot Inside
Noraalized by Total
For regression all variables on this page
43.
45.
Patch Outside Clothing
30. Head Outside
Neck Front Outside
Neck Back Outside
Upper Ara Outside
Shoulder Outside
Chest Outside
Back Outside
Foreara Outside
Hip Outside
Thigh Outside
Shin Outside
Calf Outside
Ankle Outside
Foot Outside
32.
34.
36.
38.
40.
42.
44.
46.
48.
50.
52.
54.
56.
Lbs AI
are dependent.
Variable Selection Coapleted - to VAR LIST ENTRY SCREEN
FIGURE 20. Statistics Variable List Entry Screen (continued)
2-56
-------
STATISTICS: Variable Selection List
Applicator File
Maximum variables/pathway :
Patch Inside Clothing
57. Right Hand Inside
59. Left Hand Inside
61. Both Hands Inside
63. Dernal Quant. Limit
Page 4 of 4
., 6; Reg., 24; Univ., 6
Patch Outside Clothing
58. Right Hand Outside
60. Left Hand Outside
62. Both Hands Outside
64. Hand Quant. Limit
Normalized by Total Lbs AI
For regression all variables on this page are dependent except 163 and 164.
Variable Selection Completed - to VAR LIST ENTRY SCREEN
FIGURE 20. Statistics Variable list Entry Screen (Continued)
2-57
-------
1. Time;
2. Total Lbs AI;
3. Total Lbs AI/Time;
4. Rate Lbs AI/Acre/Time; or
5. None.
If no normalization is desired, select none. Choosing "Time- will activate another menu:
1. Dermal Sample Time and
2. Air Sample Time,
where the user may select Dermal Sample Time or Air Sample .Time.
Variable selection is performed at the "Variable Selection list" screen by moving the
cursor to the appropriate variable name and pressing the ENTER key. This causes a highlight^
window to appear in which the user may specify the statistical procedures desired for the
selected variable. The user may press "C" for correlation, "R" for regression, "U" for
univariate statistics, 'X" for creation of an ASCHfile, or any combination of these. Todeselect
a procedure, the user presses the corresponding key a second time. When all procedures have
been specified for a variable, the user presses ENTER once again and may then select another
variable. To deselect all statistical procedures for a variable, the user may move the cursor to
the variable name and press the DELETE key. PHED statistical procedures are discussed in
detail in Section 3.
If the user specifies dermal dosimeter data (patches) for statistical analyses, a pop-up
window will appear on the screen for the user to specify whether statistics are to be performed
using only right patches, only left patches, or some combination of these (Figure 21). Then,
if inside clothing patches have been specified for statistical analyses, another pop-up window will
appear on the screen (Figure 22). The user will then specify whether statistics are to be
performed using inside personal clothing patches, inside protective clothing patches, inside
2-58
-------
STATISTICS: Variable Selection List
Applicator File
Maximum variables/pathway : Corr., 6; Reg., 24; Univ./?
Patch Inside Clothing
29. Head Inside
31. Neck Front Inside
33. Neck Back Inside
35. Upper An Inside
37. Shoulder Inside
CRU 39. Chest Inside
41. Back Inside
43. Foream Inside
45. Hip Inside
47. Thigh Inside
49. Shin Inside
51. Calf Inside
53. Ankle Inside
55. Foot Inside
Normalized
For regression all variables on this page are dependent.
Variable Selection Completed - to VAR LIST ENTRY SCREEN
CHOOSE FROM THE FOLLOWING POSITIONS
Use left patch
Use right patch
Use an average of the
left and right patch
Left patch and ave
Right patch and ave
Left and right patch
Left and right patch and ave
FIGURE 21. Patch Combination Selection Box
2-59
-------
STATISTICS: Variable Selection List
Applicator File
Page 3 of 4
Maximum variables/pathway : Corr., 6; Reg., 24; Univ., 6
Patch Inside Clothing
29. Head Inside
31. Neck Front Inside
33. Neck Back Inside
35. Upper Am Inside
Shonlder Inside
Chest Inside
41. Back Inside
Forearm Inside
Hip Inside
47. Thigh Inside
49. Shin Inside
Calf Inside
Ankle Inside
Foot Inside
Normalized
For regression all variables on this
37.
CRU 39.
43.
45.
51.
53.
55.
CHOOSE FROM THE FOLLOWING LOCATIONS
(LOCATION CODE IN PARENTHESES)
Inside personal clothing
(1)
Inside protective clothing
(2)
Inside personal and
(3)
protective clothing
(1) or (2)
(1) or (3)
(2) or (3)
(1) or (2) or
(3)
page are dependent.
Variable Selection Coapleted - to VAR LIST ENTRY SCREEN
FIGURE 22. Inside Patch Selection Box
2-60
-------
personal and protective dotting patches (based on Inside Patch Location values), or some
combination of these.
At any time during the variable selection process, the user may PRESS the "ESC" key
to return to the "Variable list Entry Screen" for further instructions or for clarification of the
variable selection procedure. The variables already selected win have been stored and win not
have to be reselected unless the user chooses Option 3 to clear the variables.
Once all the variables have been selected for any run, the user should press the
key to initiate the statistical procedures. Because of space limitations on the Data Report
Screens, only six variables may be selected for univariate and correlation analysis. Five
independent and five dependent variables may be selected for regression. Please note that each
patch variable may represent up to three variables; the left patch, the right patch, and an average
of the left and right patch. A screen comes up to show the user that the command is
bemg executed, and that the data will appear. Do not press any keys until all the statistics have
been calculated, and the screen prompts appear on the Statistics Reporting Screen.
The types of statistical analyses available include correlation, regression, and univariate
statistics. (The first two require that the user select at least two variables.) Due to the
maximum precision of four decimal places in all Revelation* calculation operations, there may
be a small percentage error carried over among the various computations in the Statistical
Pathway. If the data item is too large for its respective field on the output display screen, the
field is filled with asterisks.
2.4.3.1 Correlation
Each variable selected for correlation will be correlated against every other variable that
the user has selected for correlation, but because of space on the Data Report screen, only six
variables may be correlated during any one pass through the Statistics Path. The correlation
program will run only if two or more variables have been selected for correlation. The
2-61
-------
Correlation Data Report Screen (Figure 23) lists each variable by name as a separate column
across the top of the screen, then lists these same variables by number as rows along the left side
of the screen. For each pair of variables, the correlation coefficient is listed. Each correlation
coefficient value is also tested against a t-test table, and those correlation values with an asterisk
next to them are statistically significant at the 95 percent level. Below each correlation
coefficient value, the sample size of values correlated, "n", is in parentheses.
If there are data pairs in any record for which one of the values is missing, the
correlation program will delete that record from that specific correlation analysis run. If ihere
are no data pairs for any of the records, this will appear on the screen as "No Data" within that
location in the matrix. Diagonally across the screen (upper left to lower right), each variable
is correlated against itself, so these values in the diagonal are always "1.0000", indicating
perfect direct correlation. If all of the values for one variable in the data pair are identical for
aU the records, the correlation computation equations will be dividing by zero. This is indicated
by a display of "####."
Some exposure measurement data (dependent variables) win be entered and stored in the
PHED as "ND," meaning "not detected." Before any "ND" data are processed in the Statistics
or Exposure Pathways, these data are converted within their respective record to one-half the
appropriate limit of quantification for that specific record.
2-4.3.2 Regression
The Regression statistical package in PHED is a simple, linear regression equation
program. The instructions at the bottom of each variable list indicate which variables should
be independent variables and which should be dependent. The field exposure test rnnHiH™. ^
laboratory quantification limits are the independent variables, and the exposure m
(patches, air amounts, hand values) are dependent variable.
2-62
-------
CORRELATION COEFFICIENT MATRIX-SUBSET : OPEN.EC.MLOD
Normalized by Total Lbs AI
1.
2.
3.
4.
Total
Gal.
Mixed
1.
1.0000
.4368*
(118)
-.1223
(26)
-.1420
(26)
Total
Lbs_AI
Mixed
2.
.4368*
(118)
1.0000
-.1617
(26)
-.1595
(26)
Right
Hand
Outside
3.
-.1223
(26)
-.1617
(26)
1.0000
.9466*
(26)
Left
Hand
Outside
4.
-.1420
(26)
-.1595
(26)
.9466*
(26)
1.0000
Page 1 of 2 Press for next page, to exit
*** Press 'P' to print this page ***
FIGURE 23. Example Correlation Screen
2-63
-------
At least one independent and one dependent variable must be selected before the
regression program will run. No more than five independent and five dependent variables may
be selected for regression during any one pass through the "Variable Selection List". If three
of the six are independent variables and three are dependent, then there will be nine pairs of
independent-dependent variables; all other combinations will produce less than nine pairs. Iffor
any one of the records in the subset, one independent or dependent variable is missing, this
individual record will not be passed to the Regression subroutine.
The Regression Data Report Screen (Figure 24) lists (1) each of the independent
variable-dependent variable pairs that have been regressed as the left two columns; (2) the
sample size for those records (pairs of values which both had values); and (3) the slope,
y-intercept, and the coefficient of determination (which is the square of the coefficient of
correlation). Because of the similarity in the computation procedures, the correlation routine
of PHED is used to compute the regression values. This coefficient of determination value will
be a number between 0 and 1 and is an indicator of goodness-of-fit, with 1 being the best-fit and
0 the worst. The value must be used in conjunction with the sample size to obtain a
statistically-derived goodness-of-fit determination.
2.4.3.3 Univariate Statitf
The Univariate Statistics Option executes a series of statistical procedures. For each
variable selected the univariate statistics include: sample size; mean; median; standard
deviation; 10th, 25th, 50th(median), 75th, and 90th percentile; minimum value; maximum value;
range; and the lower and upper 95 percent confidence intervals (Figure 25). Because of the
space available on the Data Report Screen for univariate statistics, only six variables may be
included in any one pass through the Statistics Path.
2-64
-------
REGRESSION ANALYSIS: Mixer Loader Pile SUBSET: OPEN.EC.MLOD
INDEPENDENT DEPENDENT SAMPLE SLOPE INTERCEPT COEFF.
VARIABLE VARIABLE SIZE OF DET
(NORMALIZE CODE) (INSIDE L/R)
Total Gal. Mixed Right Hand Outside 26 -.0027 7886.2577 .0149
(N/A) (N/A N/A)
Total Gal. Mixed Left Hand Outside 26 -l0076 20062.5242 .0202
(N/A) (N/A N/A)
NORMALIZED BY Total Lbs AI
*** Press any key to continue ***
HGUBE 24. Regression Screen
2-65
-------
SUMMARY STATISTICS: Mixer Loader File
Total Total Right
Gal. Lbs_AI Hand
Mixed Mixed Outside
INSIDE CODE: N/A N/A N/A
CONFIG : N/A N/A N/A
NORMALIZED BY Total Lbs AI
SAMPLE SIZE
ARITH. MEAN
GEO. MEAN
MEDIAN
STD. DEV
PERCENTILES
10 TH
25 TH
75 TH
90 TH
VARIABILITY
MINIMUM
MAXIMUM
RANGE
95% CI LOWER
95* CI UPPER
SUBSET:
Left
Hand
Outside
N/A
N/A
26
OPEN.EC.MLOD
118 118 26
487.2542 29.8477 7884.0063 20056.144
216.1555 6.3384 246.4957 318.3834
150 12.85 1333.0357 1551.5873
804.9033 65.9974 23950.531 58440.242
50 .387 .5843 .6316
100 1.02 2.6104 2.3479
320.75 23 4758.7617 5100
1350 48.8 9130.3824 29718.549
50 .1 .1544 .1029
3920 392 121804.51 266165.41
3870 391.9 121804.35 266165.31
342.0234 17.9396 -1789.641 -3547.938
632.4851 41.7558 17557.654 43660.227
Page 1 of 1 Press to exit
*** Press 'P' to print this page ***
FIGURE 25. Example Univariate Screen
*
2-66
-------
2.4.3.4 Statistics for "Complete" Data Sets Only
If the user selects option 2 from the Statistics Variable Selection list menu, PHED win
recalculate univariate and correlation statistics for "total" body exposures from "complete" data
sets only. Further explanation of the concept of "complete" data sets is warranted. Because
most of the records in PHED are incomplete in a variable way (i.e., some records have exposure
data only for one set of body locations and other records have exposure data only for a different
set of body locations), a given subset of records may have a different number of data points for
different body locations. This fact prevents accurate development of certain types of statistics
on such a data subset. One way to circumvent this problem is to select a data set composed of
records that have exposure values for each of the body parts that are key to a given type of
operation. For example, since hand, forearm, chest, and thigh exposure typically account for
the great majority of total dermal exposure for mixing/loading operations, these four body parts
could be taken as constituting a "complete" data set for this operation. PHED allows the user
to designate the body parts that constitute a "complete" data set; no default set of body part
designations is currently built into PHED.
As before, a screen will appear asking which method of normalization to use, if any. The
user should indicate this as before. Next, the exposure scenario must be re-selected for
developing the total exposure statistics; the screen shown in Figure 15 will appear, and the user
will go through the same series of screens as before. Next, the screen shown in Figure 26 (an
example of which is shown for the Mixer/Loader file) will be displayed to allow the user to
select the criteria for what constitutes a "complete" set of body part exposures by highlighting
the desired body parts and pressing X. What will appear next is the Variable selection List that
was used before to select variables and statistical operations. Once the variables and statistics
are selected, the F3 key can be pressed to initiate processing of the total body statistics. NOTE:
DO NOT PRESS ANY ADDITIONAL KEYS AT THIS POINT UNTIL AFTER ALL THE
STATISTICS HAVE BEEN CALCULATED AND APPEAR SEQUENTIALLY ON THE
SCREEN.
2-67
-------
—Deraal StatisticsMB^BMM.
"Criteria Selection for 'COMPLETE' Data Sets
Mixer/Loader File
BEAU
NECK FRONT
NECK BACK
UPPER ARMS/SHOULDERS
CHEST
BACK
X FOREARM
HIPS/THIGHS
SHINS/CALVES/ANKLES
FEET
X HANDS
Selections are Bade by highlighting desired body part and pressing
* Select body parts constituting essentially coaplete data sets.
to continue to exit—a—^—
FIGURE 26. Complete Data Set Criteria Selection
2-68
-------
TTie output screens, which include univariate and correlation statistics, axe shown far
examples in Figures 27 and 28, respectively. The univariate output screen far "total" body
exposures is the same format as for the univariate output screen previously,
that the exposure values and statistics are expressed as pg per body part rather than jig/cm*.
The output screen for the correlation analysis deserves some explanation. Two correleation
statistics, both of which are measures of the strength of association between "total" exposure and
a given variable, are presented - the Spearman's Rank Correlations Coefficient and the
Pearson s Correlation Coefficient. These coefficients appear under the variable that is attww^
to be correlated to "total" body exposure. Hie value in parenthesis under each coefficient is the
number of data pairs on which it is based. The closer the absolute value of each statistic is to
1, the more highly associated the variable is to "total" esqmsure;
When one performs statistics on a two-variable data set, a question that frequently arises
is "how strongly are the two variables associated?" The most common definition of the term
"association" is linear association. In this case, perfect association would that:
Y = a + bX
If all data points lie on this line, the value of Pearson's correlation coefficient (R) is 1 Of b is
less than zero, R = -1), which is the strongest level of association. If the points deviate from
the line, the value of R and the strength of the association diminish. If no linear relationship
exists, then no association exists, and R - O. In Spearman's correlation, X and Y are ranked
from smallest to largest, and the degree of concordance of the ranks is If XY
pair has exactly the same rank, the Spearman's correlation = 1 (if ranks are exactly inversely
associated, then ® -1). As the degree of rank concordance diminish^ the absolute value of the
Spearman's correlation coefficient decreases. When no rank concordance exists, it equals 0.
The difference between these two measures of associate is illustrated by Figure 29. In
this example, X takes on the values 1 to 20. The Y values are generated as Y = X, Y =
XV20, Y = EXP(X)/24,000,000, and Y = RND*20, where RND is a zero-one random number.
2-69
-------
SUMMARY STATISTICS! Mixer Loader File
INSIDE CODE:
CONFIG :
Total
Gal.
Mixed
N/A
N/A
NORMALIZED BY Total Lbs AI
Total
Lbs_AI
Mixed
N/A
N/A
SAMPLE SIZE
ARITH. MEAN
GEO. MEAN
MEDIAN
STD. DEV
PERCENTILES
10 TH
25 TH
75 TH
90 TH
VARIABILITY
MINIMUM
MAXIMUM
RANGE
95* CI LOWER
50
794.36
425.9563
600
766.9743
50
103.5
1000
1800
50
3300
3250
581.7656
95% CI UPPER 1006.9544
53
52.4199
14.2486
11
79.3504
2
3.75
80
171.6
1
300
299
31.0566
73.7831
Air
Amt
of_Ch«m
~ N/A
N/A
53
3.532
.0249
.015
23.921
.0021
.0052
.0742
.2464
.0005
174.24
174.2395
-2.9082
9.9721
SUBSET:
Ave
Flow
Rate
N/A
N/A
53
166.0176
3.0906
1.175
513.441
1
1.1
5
5
.203
1840
1839.797
27.7856
304.2497
MIXLD.FILE
Total
Dermal
Exp.
N/A
N/A
53
39.8435
7.3716
10.8408
127.5345
.3784
2.091
16.6032
51.0642
.0401
811.5042
811.464
5.5078
74.1792
Page 1 of 2 Press for next page, to exit
*** press 'P* to print this page ***
FIGURE 27. Example Total Body Univariate Screen
2-70
-------
SUMMARY STATISTICS: Mixer Loader Pile SUBSET: MIXLD.FILE
Total Total Air Ave
Gal. Lbs_AI Amt Flow
Mixed Mixed of_Chem Rate
INSIDE CODE: N/A N/A N/A N/A
CONFIG ! "/A N/A N/A N/A
NORMALIZED BY Total Lbs AI
Spearman's
Rank Corr. -.1931 -.4589 .5871 -.1562
Coefficient (50) (53) (53) (53)
Pearson's
Correlation -.2147 -.1137 -.0214 .6462
Coefficient (50) (53) (53) (53)
COMPLETE BODY : FOREARMS, HANDS
Scenario : Protective overall over long pants, short sleeves, gloves
—— Page 2 of 2 Press for previous page, to exit
*** Press *P' to print this page ***
FIGURE 28. Example Total Body Correlation Screen
2-71
-------
• f
FIGURE 29. Pearson's Condatioo Versus Spearman's r^rwiatinn
. v
2-72
-------
X2/20, Y =-EXP(X)/24,000,000, and Y = RND*20, where RND-is a zero-one random
number. The respective Pearson correlations of these pairs are 1.0,0.97, 0.54, and -0.26. The
Spearman rank-correlations are 1.0, 1.0,1.0, and -0.33. Thus, in the first case, the correlation
shrinks as the association becomes less linear. In the second, the rank concordance is perfect
for aU but the random series, and this is reflected by the three Spearman rho values of unity.
Note however, that for the truly uncorrelated random case, the measures are in good agreement.
This brief exercise illustrates what the two measures of association really show. The examples
are a little contrived in that data sets rarely exhibit perfect association, either linear or
non-linear. Still, they do underscore that the two measures do respond to slightly different
features of the data.
2.4.3.5 ASCT File Creation
The user of PHED may wish to export the data out of Revelation* to plot graphics or to
perform additional statistical procedures not included in PHED. There is facility for this by
creating an ASCII file, executed by selecting a variable with the "X" key (for exporting the data)
during the variable selection procedure. This ASCII file could be exported to other PC-based
software packages. To access an ASCII file created by the statistics option, exit PHED to DOS.
The file is in the REV directory under the name specified by the user at the time of creation.
The data is set up in columns with headings for each variable. For each pass through the
Statistical Pathway, up to 24 variables may be specified for this ASCII file for data export.
2.4.4 Browse/Print
The browse option allows the user to select a data file or subset, then to select certain
variables for viewing on-screen and/or printing. The user may select as many variables as
desired, but the more variables selected the slower the processing time for PHED to format the
data for display. In addition, the display may be structured by sorting on one of the variables,
either in an ascending or in a descending order.
2-73
-------
All thc-variables for a record are formatted along the row to the right (horizontally across
the data matrix). Each subsequent record is a new row, with each record formatted vertically
down the matrix. Whenever the user moves to a new screen or the print option moves to a new
page, the record is designated in the left-hand column including die EPA Study Code Number,
the Worker Identification, and the Work Cycle Number.
When the user chooses "4. Browse/Print" on the Subset Operations menu (Figure 7), the
first of a four-screen series appears (Figure 30). This series of screens is for variable selertion
and is formatted similarly to the Subset Specification Variable Selection Menu. The cursor is
moved with the up and down and left and right arrows, and additional screens are viewed by
pressing the page up and page down keys. To select a variable, the cursor is placed on the
variable, and the user presses either the ENTER or the -+'- key. To deselect a variable, the
user presses ENTER or "+" a second time.
To select a variable for sorting the records, place the cursor on that variable and press
the S key. When the variable for sorting has been selected, a sort menu will appear in the
middle of the screen (Figure 31). Use the up and down arrow key to move the cursor and press
ENTER to specify the sort option:
ascending: sort the variables in an ascending order using the current variable as the
sort key.
descending: sort the variables in a descending order using the current variable as the
sort key.
cancel sort: do not use the current variable as the sort key and revert to the previous
sort specification, either no variable as yet or the previously specified sort
key variable.
After the user has made this selection for sorting, the user may continue selection of
variables. If the user proceeds to select another variable for sorting, the menu win indicate this
2-74
-------
BROWSE/PRINT FACILITY: Variable Selection - Applicator File
Study Information
Study Code
Worker Id
Work Cycle
Airborne Grade
Dermal Grade Uncovered
Dermal Grade Covered
Hand Grade
Sampling Date
Time Study Began
Time Study Ended
Compound Identification
Pesticide Action
Liquid (lb ai/gal)
Liquid Type
************
************
Solid (% ai by weight)
Solid Type
Page 1 of 4
Site Data
Study Lpcation
Indoor
Outdoor
Crop
Crop Height
Row Spacing
Help
Show specification
Done
Select/unselect
Exit
Sort
FIGURE 30. Browse/Print Variable Selection
2-75
-------
BROWSE/PRINT FACILITY: Variable Selection - Applicator File
Mixer/Loader Info
************
************
************
************
************
************
************
************
************
Application Info
Application Method
Rate (lb ai/acre)
Rate (gal/acre)
Total lb ai Applied
Total I Acres Treated
Final Mix (lb ai/gal)
Total I gal Sprayed
Vehicle Make & Model
I Tank Applications
Cab Type
Speed
Page 2 of 4
AI - cont.
Applicator Make/Model
Disc I
I Nozzles
Nozzle Type
Nozzle Model/No
Nozzle Pressure
Boom Height
Svath Width
Boon Length
Shank/Prob Type
Depth of Inj/Incorp
Help
•F2> Show specification
Done
Select/unselect
Exit
Sort
FIGURE 30. Browse/Print Variable Selection (continued)
2-76
-------
BROWSE/PRINT FACILITY: Variable Selection - Applicator File
< RH. HI >
Weather Data
High Humidity
Low Humidity
High Temperature
Low Temperature
Wind Speed
************
************
Wind Direction
Personal Info
Study Participant
Years Experience
Page 3 of 4
Airborne Exp Data
Technique
Air Sampling Time
Air Quant. Limit
Amount Collected
Volume Collected
Initial Flow Rate
Final Flov Rate
Average Flov Rate
Help Done :F9> Exit
Show specification Select/unselect Sort
FIGURE 30. Browse/Print Variable Selection (continued)
2-77
-------
BROWSE/PRINT FACILITY: Variahl* Ca,
Help
Shov specification
+
+
Left
LOC
Right
Left
0
+
0
0
+
0.
0
+
0
0
0
+
0
+
0
+
+
+
+
+
+
^ •
+
+
+
+
-r
+
+
+
+
*
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
4-
Done
Page 4 of 4
Sampling Time
Measure Method
Hand Measuring Method
Avg Dermal Quan. Limit
Avg Hand Quan. Limit
Incidental Contact
Exit
Sort
FIGURE 30. Browse/Print Variable Selection (continued)
2-78
-------
BROWSE/PRINT FACILITY: Variable Selection - Applicator File
< STUDY.CODE > SELECTED
Sorting Criteria
Study Code
Worker Id
Work Cycle
Airborne Grade
Denial Grade Uncovered
Dermal Grade Covered
Hand Grade
Sampling Date
Time Study Began
Time Study Ended
Selected Sort Variable:
Study Code
Please choose sort method:
Ascending
Descending
Cancel Request
Page 1 of 4
Study Location
Indoor
Outdoor
Crop
Crop Height
Rov Spacing
Help
Show specification
Done
Select/unselect
Exit
Sort
FIGURE 31. Browse/Print Variable Sort
2-79
-------
old sort variable, and the user will either revert to the old variable or wiltovenide with the new
sort variable; the old sort-variable will become a non-sort, but selected variable.
At the bottom of the variable selection screens, there is a display of option keys. The
displays a help screen for the specific variable that the cursor is on. The
key displays a list of all variables cunently selected, and the sort-key variable (if one is
designated) will be indicate (Figure 32). Pressing the key will terminate the variable
selection process and return the user to the Subset Operations menu. TTw ENTER key will
select or deselect the highlighted variable. The key will indicate that the variable
selection process has been completed, and processing of the variables will begin. A "Percentage
Completion- barometer will appear on the screen; a this point, do not press any keys until the
next selection menu appears.
n.e Operations Selection menu for Browse/Print then appears (Figure 33), giving the user
the choice of 1) an on-screen browse; 2) a hard-copy printing; 3) creation of an ASCII file; or
4) return to the series of Variable Selection menus (but,by choosing "4", all previously chosen
variables are nullified).
By selecting • 1" to browse on-screen, the user will view the variables set up as a series
of horizontal and vertical screens with the number of horizontal and vertical screens being
dependent on the number of variable selected and the number of records in the subset
respectively. THus, the variables are displayed as columns for as many horizontal screens (or
pages in the print option) as are necessary to list all the variables (Figure 34). nie records are
individual rows displayed downwards for as many vertical screens (or pages in the print option)
to list all the records. For each screen and page, the left most column is always the record
designation (EPA Record No., Worker ID, and Work Cycle No.). To view all the variables for
a record, use the right and left arrows; to view all the records for a variable, use the up and
down arrows or the page up and page down keys. Tie -home" key will move to the first of a
set of vertical screens; and if already on the first set of screens, the home key will move to the
first (left most) horizontal screen. Tie "end" key will move to the last of a set of Vertical
2-80
-------
BROWSE/PRINT FACILITY: Selected Variable Listing
Page 1 of 1
Position Sort Variable
1 Study Code
2 Airborne Grade
3 Head - inside right patch
4 Both Hands - both hands outside
5 A Study Location
6 Liquid Type
Press
-------
Browse/Print File or Subset
(4 Variables Selected)
1. Browse selected variables
2. Print selected variables
3. Create an ASCII file
4. Return to variable selection
Press the number of your selection
F9> Return to subset operations senu
FIGURE 33. Browse/Print Operations Meng
2-82
-------
screens; and if already on the last horizontal row of screens, the "end" key will move to the last
(right most) horizontal screen. To exit from the browse option, press the key.
By selecting "2" to print the browse option, the user first is asked to specify the number
of printer characters per line, either 80 or 132. The default is 80 characters per line. Once
browse print has begun, the screen shows which page is currently being printed. The horizontal
pages (each variable for a record) are printed first with numbers in the upper righthand corner
indicating the horizontal page number and vertical page number, thai the next vertical row of
pages is printed until each of the vertical pages and subsequent pages in that row have been
printed. The user has the option of prematurely exiting the browse print option by pressing the
< ESC > key; the printer will finish printing the current page and win then terminate printing.
When the user allows the print option to go to completion, pressing any key will return the user
to the selection menu.
By selecting "3", the user can create an ASCII file of selected data. The user will be
asked to input a file name; if the file exists, the user is given the option to overwrite the existing
file. The ASCII file is formatted in columns with the first three lines making up the column
heading, the fourth line indicating the width of each column, and the fifth to the end line
containing the data. For dermal patch values, "ND" is coverted to -1; "C" is converted to -2;
and "M" is converted to -3. This will allow the data to be used as a numeric field if the file is
to be read into another program. This file is located in the REV directory. The formatted
ASCn file may be used in conjunction with a commercial statistics package or imported into a
spreadsheet for simple calculations.
If the user selects "4" to specify a new set of variables with that current data file or
subset, the old set of variables is nullified with initiation of new variable selection. To exit the
Browse/Print Option when the user is at the selection menu, the user would press the < ESC >
key.
2-83
-------
<< APPL.FILE >> (H)Page 1 (V)Page 1
Dermal
Record Study Worker Airborne Grade
Code ID No. Grade Uncovered
0448*A*01
448
A
C
C
0448*A*03
448
A
C
c
0448*A*04
448
A
C
c
0448*G*06
448
G
E
c
0448*H*07
448
H
E
c
0448*F*05
448
F
E
c
0448*F*08
448
F
E
c
0439*JK*01
439
JK
E
c
0439*CM*01
439
CM
C
c
0433*SS*06
433
SS
A
B
0433*DS*04
433
DS
A
B
0415*B*5
415
B
C
0430*BH*01
430
BH
D
B
0432*RG*01
432
RG
A
A
0432*RG*02
432
RG
A
A
0423*C*01
423
C
A
A
Total Horizontal (H)Pages 1 Total Vertical (V)Pages 15 to exit
FIGURE 34. F.xamplf Browse/Print Display
2-84
-------
2.4.5
Option "5" on the Subset Operations Menu (Figure 7) allows the user to view the
specifications used to select the current subset (Figure 35). The last criteria specification utilized
to select the subset is listed first, then each preceding specification is listed down the screen.
If the current subset has been selected as a subset of another subset, the Specifications
option screen win show the specifications for that subset operation as well as the specification
criteria for the parent subset. After viewing the subset specification criteria, the user may return
to the Subset Operations screen by pressing the key.
2.4.6 Delete
Option "6" on the Subset Operations screen allows the user to permanently delete a
subset. This action does not delete the data contained in the subset, it only eliminates the subset
name and its associated specification criteria. After selecting option "6" to delete the subset,
the user is asked to confirm deletion of the subset by typing "Y" or "N". After deletion of the
subset, pressing any key will return the user to the File/Subset Selection Menu. Note that data
files cannot be deleted, only subsets may be deleted.
2.4.7 Print File/Suhwt
Option "7" on the Subset Operation menu allows the user to print the data contained in
the current file or subset. The Print File/Subset Menu appears (Figure 36), allowing the user
to choose matrix or survey form format. Figure 37 shows a sample page printed in matrix
format, and Figure 38 shows a sample page printed in survey form format.
2-85
-------
DATA ANALYSIS SECTION: File/Subset Selection
Name: AIR.OPEN.APPL
MLAP.FILE APPL.FILE FLAG.FILE MIXLD.FILE
Name found, proceeding normally ...
<< Specifications >> Page 1 of J.
Subset Specifications for AIR.OPEN.APPL
With Application Method Equal to 1 and
With Cab Type Equal to 1
Subset originated from APPL.FILE
Use Home, End, Page Up and Page Down keys to view specifications.
- Return to Subset Operations Menu
FIGURE 35. Fxamplr Subset Specifications Display
2-86
-------
Print File/Subset Records
1. Print file/subset in matrix format
2. Print file/subset by survey forms
Press the number of your selection
Return to subset operations menu
FIGURE 36. Print File/Subset Menu
2-87
-------
; 1
• ^
01 AUG 1991 The Applicator File Horiz. Page 1
Subset 0460.APPL Vert. Page 1
» v Jermal Dermal Monitoring Monitoring
1BM Airborne Grade Grade Hand Start Time End Time
Gra«e Uncovered Covered Grade (hrs) (hrs)
0460*B*01 A A
0460*A*02 A A
0460*F*01 A A
0460*E*02 A A
0460*D*01 A A
0460*D*02 A A
0460*B*03 A A
0460*A*04 A A
0460*F*03 A A
0460*E*04 A A
0460*D*03 A A
0460*D*04 A A
0460*B*06 A A
0460*F*05 A A
0460*E*06 A A
0460*A*05 A A
A
0733
1254
A
1319
1645
A
0724
1150
A
1323
1641
A
0524
1038
A
1222
1605
A
0705
1105
A
1311
1631
A
0716
1149
A
1315
1641
A
0520
1029
A
1133
1619
A
0715
1120
A
0715
1140
A
1308
1647
A
1307
1621
FIGURE 37. Matrir TWmnf prin^
V
2-88
-------
APPLICATOR EXPOSURE SURVEY DATA (DATABASE LISTING)
PAGE 1
Page 1 of Survey Form
Study ID 0460*A*02
Grades: Airborne A
Sampling Date 06-08-88
Start Time 1319
End Time 1645
Action of Pesticide
Formulation:
Liq (lb ai/gal)
Liq type
Solid (% ai by vt)
Solid type
Location CA
Indoor
Outdoor
Crop
Height
Spacing
Dermal :Uncovered A Covered Hand A
1) Fungicide
L.
50.00
2) Drv-flovable
X
GRAPES
6.0
12.0
Application Method lj_
Rate (lb ai/acre)
Rate (gal/acre)
Total lb Applied
Total I Acres
Final Nix Cone, (lb ai/gal)
Total I Gallons
I Tank Applic. 2
Vehicle Make/Mod FORD 861 TRACTOR
Cab Type 1) Open Cab (Cockpit:)
Airblast
.1875
50.0000
2.8130
15.0000
.00375
750.0000
FIGURE 38. Example Survey Farm Format Printout
v
2-89
\
-------
2.4.8 Print-Comment
This option allows the user to print the comment field for the selected subset or file.
Print Comment reads and sorts the data by study code before asking the user if the comments
are to be viewed on the screen or sent to the printer. Reading the data may take time depending
on the size of the file or subset. DQJIQI press any keys while the data are being read; this will
cause the comment menu to display briefly and will process a comment menu option or return
to the main menu (similar to what would occur during the exposure and statistics calculation
routines).
2.4.9 List/Count
The List/Count option from the Subset Operations Menu allows the user to print or
display all the record IDs contained in a file or user-defined subset. Additionally, the number
of records in the selected file or subset will be automatically displayed at the top of the
List/Count Submenu screen. When viewing record IDs for large files or subsets, the user may
press the ENTER key to proceed to the following screens that continue the listing. If the user
wishes, the list of record IDs may be printed as directed on the screen. When ready to return
to the Subset Operations Menu, press the F9 or ESCAPE key.
2.5 Back Up/Restore of User Entered Data
This option is necessary in order to save data added to PHED by the user before updating
to another version of PHED. Also, this option will reinstall the data to PHED after it has been
updated. After selecting this option, type "1" to back up data; "2" to reinstall data; or "3" to
exit. This is presented to the user as a separate screen. (See Figure 39.) When backing up
data, PHED stores the data in the DOS directory \REV\USERBACK. If the user deletes the
current version of PHED in order to load a future update of PHED, the user must NOT delete
this directory or the PHED data files will be lost.
2-90
-------
Backup/restore PHED Data Records
1. Backup PHED data records
2. Restore PHED data records
Press the nuaber of your selection
Return to PHED Main Menu
FIGURE 39. Backup/Restore Menu
2-91
-------
1
-------
3. STATISTICS DOCUMENTATION
The purpose of this section is to present detailed technical, scientific, and
documentation for the various parts of the Pesticide Handlers Exposure Database (PHED).
Sections 1 and 2 of the User's Manual are intended to provide the user with enough information
to operate the database and to understand its general aspects. This section presents detailed
narrative to support the user who desires to thoroughly evaluate the details of the database.
In the following sections, various technical aspects will be described, including material
concerning the inhalation and dermal exposure calculation algorithms and the
calculations.
3.1 Exposure Pathway
Since the 1960s, scientists have been performing field studies to estimate the amount of
exposure received by workers applying pesticides. Dosimeters comprised of pads and patches
of various material (including the actual fabric of workers' clothing) have been placed on various
body part locations during various work cycles for subsequent extraction and analysis in
analytical laboratories. Until recently, standardized protocols for conducting exposure studio
did not exist, however, data have been developed which, when taken as a whole, could assist
in the decision-making process concerning worker exposure. Thus, the PHED has been
developed as a repository for quality exposure data, as well as a software system to perform
exposure computations.
The system is designed to be a generic exposure database. This means no information
is included on the specific active ingredient(s) or formulation names used in any study.
Additionally,, each replicate or record from each different study is treated equally as a composite
of different body parts. Although these studies were conducted to sample different body parts,
exposure is not computed on a record by record basis because very few of the available studies
have evaluated the exposure to all body parts. Instead, PHED evaluates exposure to each body
3-1
-------
pan across all the records in a data file or subset to develop an exposure-estimate for each body
part, then exposures to each body part are added to develop a composite tor total
dermal exposure.
Data on exposure to each body part are entered in the database in units of fig/cm2.
Exposure computations involve calculating the mean, median, and geometric mean for each
record having data for that body part, then multiplying the respective values times the body part
surface area value to obtain exposures in mass units Gig). For hands and inhalation exposure
estimates, the values are entered in the database in mass units: the amount measured on each
hand, the total amount measured on both hands, and the amount collected by the inhalation
sampling system.
The exposure calculations are programmed to be normalized based on one of the
following: 1) time; 2) total pounds (lbs) active ingredient (AI); 3) total lbs Al/time;
4) application rate (lbs Al/acre/time); and 5) none. Each normalization is performed on a
per-record basis. The estimates are normalized and composited for all body parts for all the
records in the data file or subset. The normalized results will be displayed on the screen.
3.1.1 Inhalation Exposure Algorithm*
The Inhalation Exposure Algorithms are programmed to compute exposure for each
record. The following are the steps taken in calculating inhalation exposure:
(1) For each record in the file or subset:
(a) Obtain value for normalization by either 1) time; 2) total lbs AI; 3) total lbs
Al/time; 4) rate lbs AI/acre/time; or 5) nmy
(b) Obtain values for:
• airborne detection limit (AIR.UMIT (/tg));
• amount of airborne chemical collected (AIR. AMOUNT (ug)V and
• average flow rate (AIR. AVE (liters/min)).
3-2
-------
Ifthe amount of airborne chemical collected is not detectable, then
me value assigned is one-half of the quantification limit:
AIR. AMOUNT = AIR.LIMIT/2.
(c) Calculate the normalized exposure for each record:
® ^ personal air pumps or air impingers were used
(AIR.TECH.CODE = 2 or 3), then
Exposure = Ant^OUNT * INHALATION RATE Ohershmr*
AIRAVE * NORMAUZATIONVALUE
[If air.AVE value is missing, then record is skipped.]
(ii) If a filter or respirator was used
(AIR.TECH.CODE = 1), then:
AHLAMOUNT
NORMALIZATION VALUE
(d) Ifthe inhalation exposure has a non-zero value for that record, the value is saved
m an array.
(2) For the exposure array, calculate median, average, variance, and standard deviation.
(The equations are identical to those used in the statist^ pathway.)
(3) Calculate exposure confidence intervals, as follows:
(a) exposure (normalized /xg)
Lower limit = average - (1.96 * standard deviation)
Upper limit = average + (1.96 * standard deviation)
(b) exposure fig/Kg active ingredient (when selected by user within option for
normalizing by amount active ingredient)
3-3
-------
Lower limit = (average - (1.96 « standard deviation)) *&453
Upper limit = (average + (1.96 * standard deviation)) * 0.453
3.1.2 Dermal Exposure Algorithms
The dermal exposure algorithms build arrays of data and perform computations using the
appropriate values found on the next-to-last and last data entry screens, which correspond to the
last page of each set of Exposure Survey Forms. If the dermal exposure algorithm utilizes
outside patches, these data are found on the right half of the data entry screens and the right half
of the Exposure Survey Forms. If the dermal exposure algorithm utiiiw* patches, the
data are on the left half of the screen and left half of the Survey" Form. For patches, the
Location Code is the key discriminator in determining which patch values are used and which
patch values are not used in each of the various exposure scenarios.
The following are the steps taken in determining dermal exposure:
(1) For each record in file or subset:
(a) Obtain values for normalization based on variable selected by the user. If the
normalization value is missing, then this record is ignored.
(b) For each body location in the data file, determine whether the inside patch is
appropriate. This is done by comparing the location code for each body part with
a scenario-specific code. (See Table 2.) If the location code is not appropriate
then that patch value is not used.
(c) Process all variables with non-numeric values (those with values of "ND", "C",
or "M"). (See Table 3.) Variables with a value of "ND" are replaced by the
dermal quantification limit value divided by two. Variables whose values are "M"
are set to zero for that side of the body, and the companion value is used (the
value reported for the other side of the body). Values of "C" are set to zero for
that side of the body, and the comparison value is divided by 2.
(d) Assign surrogate exposure values. (See Table 4.)
(e) Normalize exposure values by the selected variable. If this variable has a missing
value, then the exposure data for that normalization factor is discarded.
3-4
-------
Table 2. M^riiT ***nt Each Dannal
and the Applicable Patch Location Codes
Ssaaoa.
Patch
1
2A
2B
2C
2D
3A
3B
3C
3D
3E
3F
Head
E
E
E
E
E
E
E
E
E
E
E
Neck Front
E
E
E
E
E
E
E
E
E
E
E
Neck Back
E
E
E
E
E
E
E
E
E
E
E
Upper Ann
E
1
1
1
1
3
3
3
3
2
2
Shoulder
E
1
1
1
1
3
3
3
3
2
2
Cheat
E
1
1
1
1
3
3
3
3
2
2
Back
E
1
1
1
1
3
3
3
3
2
2
Foreerm
E
1
1
2
2
3
3
2
2
Hip
E
1
1
3
3
3
3
2
2
Thigh
E
1
1
1
1
3
9
• 3
3
2
2
Shin
E
1
1
1
3
3
3
3
2
2
Calf
E
1
1
1
1
3
3
3
3
2
2
Ankle
E
1
1
1
3
3
3
3
2
2
Foot
E
I
I
I
I
I
I
I
I
I
I
Hande
E
E
I
E
I
E
I
E
I
E
I
1 - Uae only internal patch vainea with a location code of 1.
2 - Uac only internal patch vafawa with • location code of 2.
3 - Uae only internal patch valuaa with a location code of 3.
£ - Uae only external patch vainea.
I - Uae only inaide patch vainea.
Scenario:
1. No clothing.
2. Noonal work clothing.
A. Long panta, abort aleevea, no glovea.
B. Long panta, rikoit aleevea, glovea.
C. Lang panta, long aleevea, no glovea.
D. Long panta, long aleevea, glovea.
3. Protective clothing.
A. Protective ovecall over long panta, rfiort aleevea, no glovea.
B. Protective ovecall over long panta, Ant aleevea, glovea.
C. Protective ovecall over long panta, long aleevea, no glovea.
D. Protective overall over long panta, long aleevea, glovea.
E. Protective overall over no clothing, no glovea.
F. Protective overall over no, clothing, glovea.
3-5
-------
TaMe.3. Procedure for Assigning Numeric Values to Exposure Fields witWioo-Numeric Data
SINGLE PATCH
Present
Missing
ND
ND
POL
Present
Missing
STORED VAT TTP
Single Pitch Value
DQL/2
T.EFT patch
Present
Present
ND
ND
•M'
"M"
Present
Missing
¦C"
"C"
Present
Missing
RIGHT PATCH
Present
Mining
POL
STORED VALUES
Missing
ND
ND
Present
Missing
•M"
"M-
Present
Missing
"C*
"C"
LEFT
RIGHT
AVERAGE
Left
Right
(L+R)/2
Left
—
Left
Right
Right
DQL/2
DQL/2
(L+RV2
—
—
Right
—
—
Left
—
—
—
—
Right/2
'
—
Left/2
Method of handling non-numeric patch values for yul analysis.
Missing - data are not entered in this field in the data file.
N/A - not applicable.
DQL - Dermal Quantification Timft
ND - Non Detectable.
"M" - Middle patch.
"C" - Combined value.
3-6
-------
Table 4. Specific Body, Patch, and Surrogate Locations-
Uaed in die PHED Expocore Algorithm
Body location
Head
Neck front
Neck back
Upper anas
Chest
Back
Forearm
Upper leg
Lower leg
Foot
Handb
Patches used*
Head
Neck front
Neck bade
Shoulder, upper arm
(W
Back
Forearm
Hip, thigh
Shin, calf, ankle
Foot
Hand or both hands
Surrogatea oaed
Back, cheat, shoulder
Cheat
Back
None
None
None
None
None
None
None
None
For body parts with left and right patches (excluding hands* the value fer rfrill .
the exposure value for the hands is twice the value of the non-missing left or right patch.
3-7
-------
(2) For each body location:
(a) Calculate the median, average, and coefficient of variation. (See statistical
equations for method.)
(b) Calculate, the geometric mean and determine how data are distributed. (See
discussion of frequency distributions in paragraph 3.1.3.)
(c) Multiply the median, mean, and geometric mean by the body location surface
area. (See Table 5.)
(3) For composite values (total dermal exposure):
Exposure data from each body part are added together to determine the composite
exposure, the total dermal exposure based on that scenario for the data records in that subset or
data file.
(a) For the column headed with "Distrib. Type," add the values from respective
columns, using the "mean" value if there is a normal distribution; using the
^geometric mean" value if there is a lognormal distribution; an
-------
Table 5. Suifrce Aims Uaed for Expoeore Eatuutian
Head
Eatimted
Actual
Both
Neck front
Neck back
Upper ran
Cheat
Back
Forearm
Thigh
Lower leg
Feet
Haade
1300 cm2
1300 cm2
1300 cm2
150 cm?
110 cm2
2910 cm2
3550 cm2
3550 cm2
1210 cm?
3820 cm2
2380 cm2
1310 cni2
1®
* No conversion needed, hand valnea already in microgiaaH.
-------
frequency distribution); or do not fit either a normal or a lognormal distribution (known as die
Kolomogrov-Smirnov Test). The following steps are taken in testing the distribution type of a
set of values:
(1) For each value, (input as array Y)
Determine the rank of Yi:
the rank of Yi is its position in the array Y, all repeated Yi's are assigned the
highest rank of the repeated Yi's.
e.g., Y = 2 4 5 7 7 10 11 11 11 15
rank =12355 6 9 9 9 10
Calculate Si of array S (the empirical distribution function).
Si = rank of Yi/number of Yi's (n)
Calculate the ln(Yi).
n
(2) SL = E ln(Yi)
i =1
n
(3) SL2 = E (ln(Yi))2
i =1
(4) Calculate mean and standard deviation of log transformed data
(a) LNMean = SL/n
(b) LNDev = ((n * SL2) - (SL2))/n(n-l)
(5) Calculate the test statistic.
For each Yi:
(a)
(b)
(c)
3-10
«
-------
(a) The standardized Xi is calculated as follows:
Xi = (Yi - mean)/st. dev.; if st dev. = 0, thenXiis
assumed to be 0 and the
frequency distribution is
classified as "other."
(b) The theoretical distribution function of the normal distribution is approximated as
follows:
Ti = 1/(1 + 0.2316419 * Xi)
(c) Pi = 1 - (e"Bl/2/^ * (bj * Ti + b3 * Ti3 + b4 * Ti4 + b5 * Ti5))
where
b! = 0.319381530
bj = -0.356563782
b3 = 1.781477937
b4 = -1.821255978
b5 = 1.330274429
The test statistic = the maximum of | Si - Pi | (that is, the absolute value of Si
minus Pi).
If the test statistic is greater than T (Table 6), then the test is said to "fail."
In this program, the data is first tested for a "lognormal" distribution, so the log
transformed data, LNMean, and LNDev are used to calmiat<» the test statistic.
If the test "foils," the data are not lognormally distributed, and the test statistic
is re-calculated using the untransformed data (original data). If the test "foils"
again, the data does not fit a "normal distribution" and is said to be "other." If
the test statistic is less than T, the test "passes", and the distribution is the type
being tested for; in this program, either lognormal or normal.
(6) Calculate the geometric mean of the untransformed (original) data:
lGeo. Mean =
can also be expressed in terms of its equivalent form in which the geometric mean is the
nm root of the product of the. values or:
n yj Xl . X2 ... Xm
3-11
-------
Table 6. Table of T-Statistic Values for Tasting Frequency Distributions
Number of values (n) T
1
.950
2
.776
3
.636
4
.565
5
.509
6
.468
7
.436
8
.410
9
.387
10
.369
11
.352
12
.338
13
.325
14
.314
15
.304
16
.295
17
.286
18
.279
19
.271
20
.265
21
.259
22
.253
23
.247
24
.242
25
.238
26
.233
27
.229
28
.225
29
.221
30
.218
31
.214
32
.211
33
.208
34
.205
35
.202
36
.199
37
.196
38
.194
39
.191
40
.189
->40 1.22V*
3-12
-------
3.1.4 Calculation of the Confidence Limits for Total Body Exposure
For the "arithmetic mean" total dermal value, the total body exposure is based on the
sample mean for each body part The following represents the procedure used to calculate the
lower and upper confidence limits:
Total exposure = (area of the head) . (mean of the sample for head)
+ (area of the neck front). (mean of sample)
+ ... [for all body parts: + (area of part K)(mean of sample K)]
i.e.,
Assuming that the data for each part are from a normal distribution (based on the results
of the goodness of fit test), each sample mean also follows a normal distribution. The variant*
of the sample mean (e.g., Xj is estimated by S^, where S? is the sample variance and n, is the
sample size. The variance of the total exposure is:
Var(TE) - E
i-i
Aft2
+ 2
i
-------
3.2 Statistics Pathway
3.2.1 Technical Information
Univariate statistics and correlation and regression analyses are available options in the
Statistics Pathway. A maximum of six variables may be selected for univariate and correlation
statistics and five independent and five dependent variables for regression. Variable values used
for statistical analysis are read from the user specified subset or file and processed on a record
by record basis. Value processing includes the handling of non-numeric data, normalization,
value selection based on patch configuration by body part and by right vs. left vs. average, and
value selection based on a patch location code.
The following is a general overview of the steps taken when processing variable values
on a record-by-record basis:
1. Obtain the dermal and air quantification limits*
2. For patch variables, determine if the value is for an indH* pa**-
3. For inside patch variables, obtain the location code;
4. Accept or reject an inside patch value based on location code;
5. For patch variables with left and right components, obtain the values for both
right and left patches;
6. Assign the appropriate values (Table 3) for non-numeric patch values;
7. Calculate average of left and right patch values if requested;
8. Save value in value array; and
9. Save normalization variable values in normalization array.
After all records have been processed, there exists an array for each variable
containing either numeric or missing values.
3-14
-------
• Normalize value arrays, if requested;
• Perform univariate statistics, if requested;
• Perform correlation statistics, if requested; and "
• Perform regression statistics, if requested.
3.2.2 Non-Numeric Data
Variables used in statistics may contain non-numeric data, and when this happens, the
variable is assigned a new value or the value is not used in statistical calculations. Non-numeric
values can be an "M" for middle patch value, a "C" for combined patch value, or an "ND" for
non-detectable. An explanation of these values and how they are handled in the PHED statistics
pathway is given below.
"M" " Thjfxlvaluf * ^ chest and back patch variables and represents a
middle patch. Since the chest and bade patch variables both have left and right
components, the "M" can appear in either place and signifies that the other
component contains the value of the patnh
If statistics are to be done on either the right or left patch, then the middle value
is not used. When statistics are to be performed using the average of the right
and left patches, the middle value is used as the average value.
C" - This value can be found in any patch variable that contains left and right
components. It means that two patches were exposed in the field but combined
m the lab for analysis, and only one value was determined. There is no value
for theindividual left and right patches since the patches were combined. When
the C 15 in the left patch variable, the combined value is found in the right
patch variable and divided by 2, and vice versa.
If statistics are to be done on either the right or left patch, then the combined
value is not used. When statistics are to be performed using an average value
for the right and left patches, the combined value is used as the average value.
"ND"- This value represents a non detectable value. The "ND" is replaced by the
dermal quantification limit divided by two for a dermal exposure variable or by
the air quantification limit divided by two for air exposure variable. For studies
for which a quantification limit is not reported, the detection limit divided by two
is used for "ND" values.
"ND" variable = DERM.LIMIT/2 - for dermal variables
ND variable = AIR.LIMIT/2 — for inhalation variables
3-15
-------
When the quantification limit variable contains a missing value, the variable
containing die "ND" is not used.
3.2.3 Normalization
Variables may be normalized by a user selected variable on a record-by-record basis.
A variable is normalized by dividing its value by the value of the desired normalization variable.
Normalization will occur only when the normalization variable and the variable to be nnrmaii-y^
both have non-missing values. When this is not the case (i.e., one of the value is missing),
normalization is not possible, and the exposure value is not used for statistic^ analysis.
3.2.4 Accepting or Rejecting Tnridg Patch Value
When an inside patch variable is selected for statistical analysis, the user must specify
acceptable values for the associated location code. Values for the patch variable will be used
only when the location code for that variable matches the specification indicated by the user.
3.2.5 Patch Variable Configuration
When a patch variable with left and right components is selected for statistical analysis,
the user must specify which variable configuration is desired. Any permutation of the right
patch, the left patch, or the average plus the right and/or left patches may be specified. If the
desired patch configuration for a given variable is the left patch, the right patch, and the average
of the two, then the user has selected three different variables for analysis. Since a maximum
of six variables may be selected for any one statistics option, this maximum could be reached
after selecting only two variables. When more than six variables have been selected, only the
first six will be used; the additional variables will be ignored without providing notification to
the user.
An average value is computed for statistical analyses only when there exists numeric
values for both the left and right patches. Patches with non-numeric values will have been
converted, if possible, to numeric values prior to raimiating the average value.
3-16
-------
3.2.6 Univariate Statistics
Summary statistics are calculated for the records in a subset or data file on a variable by
variable basis. AH missing values are removed from the value array prior to calculating the
summary statistics. For a listing of the equations used, see 3.2.9 below, "Steps ^ when
performing statistics on a specified variable."
3.2.7 Correlation Statistics
The correlation option will correlate each variable against all the other mrwiaHnn
variables specified by the user to produce a correlation matrix. Each pair of variables to be
correlated are processed on a xecord-by-record baas, and values for a given record are "yd for
correlation only when both variables contain numeric values. For a listing of the equations used,
see paragraph 3.2.9.
3.2.8 Regression Statistics
Regression statistics are calculated in an identical manner as correlation except
that the variables are separated into dependent and independent variable pairs.
3.2.9 Performing Statistics on a Specified Variable
(1) For each record:
*
(a) Process value and place in array X;
(b) Place value of normalization variable (TOTAL LBS AI) in array AI; and
(c) Place value of normalization variable (DERMAL SAMPLE TIME) in array
DERM.
(2) Normalize all values as requested by user.
(3) Calculate summary statistics (input is array X).
(a) Copy all non-missing values from array X to array S
(b) Sort the data in array S in ascending order
3-17
-------
£ Sj
(c) Mem = ~
n
«0
Std. dev. - ^
Eft - wttnf
n-1
(e) Lower 95 percent confidence limit on the arithnwri<»
Low « mean - T * ^ ^ (See Table 7 )
V*
(f) Upper 95 percent confidence limit on the arith™**^ m~»n.
Upper « mean + T » StA dev. ^ ^
(g) Median (50 percentile):
If n is odd, median = S( Ln/2j + 1);
if n i, ^ —«„ . Sdagp ¦> sdaol ~ n
2
where | n/x | = the next lowest whole number (i.e., 2.5 = 2)
(h) Minimum = S(l) (i.e., first element in the array of ranked data)
Maximum = S(n) (i.e., the nth or last in the array of
ranked data)
Range = S(n) - S(l)
(i) 10 percentile:
if n < 10, the 10 percentile = S(l)
if n >. 10,
low = Ln/10j
high = low + 1
10 percentile ~= Sflow) + (S(high) - Sflow)). (n/10 - low)
(j) 25 percentile:
if n < 4, then 25 percentile » S(l)
3-18
-------
Table 7. t-Statistic Table Utilized to Test for 95% Significance of
Number of
values (n) T
2
12.0760
3
4.3027
4
3.1825
5
2.7764
6
2.5706
7
2.4469
8
2.3646
9
2.3060
10
2.2622
11
2.2281
12
2.2010
13
2.1788
14
2.1604
15
2.1448
16
2.1315
17
2.1199
18
2.1098
19
2.1199
20
2.0930
21
2.0860
22
2.0796
23
2.0739
24
2.0687
25
2.0639
26
2.0595
27
2.0555
28
2.0518
29
2.0484
30
2.0452
>30
1.9600
3-19
-------
if n >. 4,
low = [n/4J
high = low + 1
25 percentile = S(low) + (S(high) - S(low)) , (n/4 - low)
(k) 75 percentile:
low = 13 * n/4j
high = low + 1
75 percentile = S(low) + (S(high) - S(Iow)) , (3 . n/4 - low)
(4)
0) 90 percentile:
low = |_9 * n/lOj
high = low + 1
90 percentile = SGow) + (S(high) - saow)) « (9 , n/10 - low)
Perform correlation on each set of paired variables whose values are represented by array
X and array Y. Both arrays must contain the same number of elements.
(a) If Xj and Yj are missing or
Xj is missing or
Yj is missing, then the variable pair from that record is
not included in the correlation analysis.
-------
(c) Pcrfoiw T test
Test variable - the absolute value of: coef *
n-2_
1-coef2
If test variable > T, then correlation is signifirgnf
Perform regression analysis on each independent/dependent variable pair represented
array I and array D:
The regression analysis is essentially a repeat of Step (4) above.
3-21
-------
-------
APPENDIX A
Data Entry Reference Tables
-------
-------
TABLE A-l. PHED User Commands for Data Entry Screens
Field prompt
Key
Function
Fl
Displays help screen for the current field
F2
Retrieves field value from the previous record
F7
Erases current field value from beginning of line to cursor position
F8
Erases current field value from cursor to end of line
ENTER
Advance to next field
CTRL/E
Cancels any changes made to the current record and exits data entry
screen
CTRI/T
Cancels any changes made to the current record and restarts entry
screen
T 4
Moves from field to field within the current record
<— -»¦
Moves cursor within the current field value
HOME
Moves cursor to beginning of current field value
END
Moves cursor to end of current field value
INSERT
Toggles between character insert mode and character overwrite mode
DELETE
Deletes character at current cursor position
BACKSPACE
Deletes character to left of current cursor position
Bottom of screen "EDIT?" prompt
Entry
Function
Displays help screen
Field number
Allows the user to edit the specified field
Cancels any changes made to the current record and exits entry
screen
Advances to next screen (if creating new record) or advances to screen
selection prompt (if editing existing record)
FILE
Saves the current record (with changes) and restarts data entry screen
Deletes current record
Cancels any changes made to the current record and restarts Hat* entry
screen
A-l
-------
TABLE A-l. FEED User Commands for Data Entry Screens (continued)
Multivalue field "WHICH VALUE?*
Value number
ADow die oser to edit the specified fidd vahm
ADowi the Mer to add a field value
dean all values for the field
Deletes a specified field value
Allows the uaer to insert a field valae at a specified location
Screen selection -PRESS FOR NEXT SCREEN, (S) FOR STARTING SCREEN, OR NUMBER?-
Screes mrmhar
Cancels any changes made to the current record and restarts data entry screen
Retrieves the specified data entry screen for the current record
*
A-2
-------
TABLE A-2. Surface Areas for Regions of the Adult Body aniMLocatioiis of Dennal
Exposure Finds that Represent These Regions
Region of the body
Surface area (cm2) of region
Location of pad(s)
representing region
Head
1,300*
Shoulder, back, chest"
Face
6S0
Chest
|
Back of neck
110
Bade
Front of neck'
150
Chest
Chest/stomach
3,550
Chest
Back
3,550
Back
B Upper arms
2,910
Shoulder and forearm/
upper arm
Forearms*
1,210
Forearm.
Hands4
820
Thighs4
3,820
Thigh
Lower legs4
2,380
Shin 1
Feet4
1,310
1
* Surface area for the head includes the 650 cm2 surface area.
b Exposure to the head may be estimated by using the mean of the shoulder, wir and chest
patches, or by using a head patch.
e Includes "V" of the chest.
4 Surface area includes left and right sides of body part.
Source: USEPA (1987) Pesticide Assessment Guidelines. Subdivision u. Applicator
Exposure Monitoring.
A-3
-------
TABLE A-3. Explanation of Coded Fields to Applicator Data Enby Scran
Action of Fntidde
1 Fungicide
2 Herbicide
3 Tnsnctiride
4 Fmmgant
5 Plant growth regulator
Liquid Formulation Type
1 EimlaifisMe concentrate
2 Aqueoua suspension
3 Microencapsulated
4 Solution
5 Undiluted liquid
Solid Fonmdatiou Type
1 WetUble - powder
2 Dry - flowable
3 Duat
4 Granule
Method of Application
1 AixUast
2 Groandboom tractor
3 Groandboom truck
4 Aerosol can
5 Aerial - fixed wing
6 Aerial - rotary wing
7 Low pressure wand
8 Paint brush
9 Backpack
10 Airless sprayer
11 Rights-of-way sprayer
12 Rights-of-way (cannon type/rail car)
13 High pressure hand wand (greenhouse &
ornamentals)
14 High pressure hand wand (Cheat-Lawn type)
15 Garden hose-end sprayers (residential uses)
16 Termiticide jnjyi ji m
17 Solid broadcast spreader (tractor/truck/AG usee)
18 Solid broadcast spreader (belly grinder)
19 Solid broadcast spreader (Scotts type residential)
20 Specialty airMast (grapes, golf course, etc.)
21 liquid broadcast spreader
22 Other
Cab Type "
1 Open cab (cockpit)
2 Closed cab (cockpit) open
3 Closed cab (cockpit) closed
4 Closed cab (cockpit) closed,
filtered air
Shank/Hrobe Type
1 Knives
2 Sweeps
3 Other
Wind Speed
1 Zero to five
2 Six to ten t*
3 Greater than ten mph
Wind Direction
1 Parallel
2 TVrp—
3 Other
Study Anticipant
1 Private applicator
2 Commercial applicator
3 Certified applicator
4 Non-certified applicator
5 Other
Applicator Yean of Experience
1 Less than one
2 One to five
3 Six to ten
4 Eleven to fifteen
5 Sixteen to twenty
6 More than twenty
Airborne Technique
1 Filter an respirator
2 air pump
3 Fixed location
Denial Measuring
1 Patch technique
2 Solvent rinse
3 Clothing
4 Other
Hind Measuring
1 Hand rinse
2 Gloves
3 Other
Incidental
1 Hand
2 Face contact
3 Other body parts
A-4
-------
TABLE A-4. Explanation of Coded Fields to Mixer/Lotder Date Entry Screen
Action of Pesticide
1 Fungicide
2 Herbicide
3 Insecticide
4 Fumigant
5 Plant growth regulator
Liquid Formulation Type
1 Emulsifiable concentrate
2 Aqueous suspension
3 Microencapsulated
4 Solution
5 Undiluted liquid
Solid Formulation Type
1 Wettable - powder
2 Dry - flowable
3 Dust
4 Granule
Packaging Type
1 Bag
2 Can
3 Bottle
4 Water soluble packet
Mixing Procedure
1 Open
2 Closed, mechanical pump
3 Closed, gravity feed
Type of DOnent
1 Water
2 Vegetable ail
3 Petroleum oil
4 Solvent
Tank Additive
0 None
1 Spreader/sticker
2 Surfactant
3 Dye
4 Other pesticide
Wind Speed
1 Zero to five mph
2 Six to ten mph
3 Greater than ten mph
Wind Direction
1 Parallel
2 Perpendicular
3 Other
Airborne Technique
1 Filter on respirator
2 Personal air pump
3 Fixed location
Dermal Measuring Method
1 Patch technique
2 Solvent rinse
3 Clothing
4 Other
Hand Measuring Method
1 Hand rinse
2 Gloves
3 Other
Incidental Contact
1 Hand contact
2 Face contact
3 Other body parts
A-5
-------
TABLE A-5^-1
1 Fungicide
2 Herbicide
3 Insecticide
4 Fumigut
5 Plant growth regulator
Liquid Fonuletioa Type
2 Aqueous suspension
3 Microencapsulated
4 Solution
5 Undiluted
Sofid Foraubtioa Type
1 Wettable - powder
2 Dry - flowable
3 Duat
4 Granule
Method at AppBcatfoa
1 Aixblast
Groundboom tractor
Grouadboom truck
Aerosol can
Aerial - fixed wing
Aerial - rotary wing
Low preaauie hand wind
Punt brush -
Airiest sprayer
Rights-of-way sprayer
Rights-of-way (Cannon type/rail ear)
High pressure hand wand (greenhouse
-------
TABLE A.-& Explanation of Coded Fields to Flagger Data Entry Screen
Action of Pesticide
1 Fungicide
2 Herbicide
3 Insecticide
4 Fumigant
5 Plant growth regulator
Liquid Formulation Type
1 Emulsifiable concentrate
2 Aqueous suspension
3 Microencapsulated
4 Solution
5 Undiluted liquid -
Solid Formulation Type
1 Wettable - powder
2 Dry - flowable
3 Dust
4 Granule
Method of Application
1 Aerial - Rotary wing
2 Aerial - Fixed wing
Wind Speed
1 Zero to five mph
2 Six to ten mph
3 Greater than ten mph
Wind Direction
1 Parallel
2 Perpendicular
3 Other
Airborne Technique
1 Filter on respirator
2 Personal air pump ,
3 Fixed location
Dermal Measuring Method
1 Patch technique
2 Solvent rinse
3 Clothing
4 Other
Hand Measuring Method
1 Hand rinse
2 Gloves
3 Other
Incidental Contact
1 Hand contact
2 Face contact
3 Other body parts
A-7
-------
-------
APPENDIX B
Exposure Survey Forms
-------
-------
Study Cod* No.:
(EPA U— Only)
Soap ling tart*
Tiaa of Day Monitoring Took PI
APPLICATOR EgQMWE SURVEY ram
(ONE FORM PER REPLICATE)
(im Military tiaa):
from To
COMPOWC IDEWTIFlCATtON "Action of Pastlelda: 1) Funglclda;
2) Harblclda; 3) Insactlclda; 4) Fmigant; 3) Plant Growth Regulator
•Formulation: Liquid (Ibal/gal) or Solid (f al by might)
1) Eaulslflabia Cencantrata
2) Suspanslon (aquaous)
3) Mlcroancapsulatad
4) Soltftlon
3) Liquid (undllotad; a.g., fwlgant)
SITE PftTft 'Location (Stata/ProwInoa and Country)
•(cfcack ooa) Indoor or Outdoor "Crop
•Crop Halght (faat)
1) Nattabla-povdar
2) Dryf loaabla
3) Oust
4) Braaula
Spacing (faat)
fiPHlfifiTlftl IHFflRfflTinN •"rfhod of Application:
Morkar
1.0.
•orkuCycia
Blank
Location o
Morkar
1.0.
Ma instn
* • Data:
Work Cycla
Noabar
1\ V A*rl?1 ~ tor«nr Wing 10) Shank Injactlon
2) Groimdbon Tractor 7) Low Prassur* Hand Hand 11) Fuarigatlon
3) Growtdboca Truck a) High Prassura tfcnd Hand 12) Solid Broadcast Spraadar
4) Groundboca Rail Car (attachad to truck) 13) Liquid Broadcast Spraadar
5) Aarlal - Flxad Wing 9) Backpack 14) othar (dascrlba)
If 14, dascrlba
Rata: (lb al/acra) and (gat/acra) Total lb al Appllad
Total t Acras Traatad PI—I Mix Concantratlon (lb al/gal Dlluant)
Total t Gallons Spray ad § Tank Applications Menltorad ;
•Vahlcla Maka and Modal:
•Cab Typa:
1) Opan Cab (Cockpit) 3) Cloaad Cab (Cockplt)/Vlndo« Cloaad
2) Cloaad Cab (Cockplt)ZWIndow Opan 4) Cloaad Cab (Cockplt)/Wlndoa Cloaad/Flltarad Air
B-l
-------
Study Cod* Mo.:
t£PA U» Only)
APPLICATOR EiPOSURE SURVEY FORM . 2
Worker
1.0.
Do Not ^
"ork_Cycl<
iosctlen
forker
see I iwt
* Data:
ttorfc^Cycla
APPLICATItH I FORMAT I ON tCan+.m Ground Spaad Caph)
•Appl icator iwu and Modal : _
"Disc /
*§ Noxzias
•Nozzia: Typa/MiiMifactarar - Modal/Mo.
•Nozzla Prassura (psl)
Height Abova 8rotnd (fact)
•Swath Width (faat) - *Langth of Boom (faat)
•Shank/Probe Typa: 1) Knlvas 2) Saaaps (3) Othar
•Oapth of Injact ion/Incorporation dachas)
Daacrlba Application Proeadnra
WEATHER DATA J Ralativa Huaidlty: High aWd Low
Tanparatura (Fahranhalt): High and Low
3) > 10 aph
Wind Spaad: 1) 0*5 aph 2) <*10 T**
Wind Dlraction Ralativa to Traval: 1) Paral lal 2) Parpandicutar 3) Othar
If 3, axplain
PERSONAL INFORMATION is Study Participant a: 1) Prlvata Applicator
2) Canarcial Appl Icator 4) Non-CarMf lad Applieator
3) Cartlflad Applicator 5) Othar
Years of Exparlanoa as an Applicator: 1) < 1 2) 1-3 3) 6-10 4) 11-15
3) 16-20 6) > 20
B-2
-------
Study Coda No.r
(EPA 0— Only)
APPLICATOR E)POSURg St* VET FORM - 3
AIH80WC EXPOSWE DATA Technique M: i) FIlT.r en Respir^r
2) Personal Air Ptap 3) Find Location
• f 3, specify location
Air Saapllng Tie# (ain) '
Aaount of Chealcsl Collected (19)
Flow Rate (L/aln): Initial
Quantification Llalt (09)
_______ Air Vol una Collected (1)
Final
. Average
Exp0SWE DATA DenMl Saapllng Tlae (hrs)
•Method of Measuring DenMl Expos ore Excluding Hands:
1) Patch Technique 2) Solvent Rinse 3) Cloth 14) Other
If 1, describe patch construction (i.e., gauze, alphe-celIulose, etc.
Location of * beta:'
Worker
1.0.
Merk CycU
ir
s— tM^n^nar
If 2, type of aolvent
if 3, describe clothing
If 4, describe
•Method of Measuring Hand Expoaura: 1) Hand Rl
If I, type of aolvent
2) Gloves 3) Other
If 2, describe gloves
If 3, describe
%
B-3
-------
Study Cod* No.:a
(EPA Uae Ontv)
APPLICATOR EXPOSURE SURVEY form - 4
Oemal Exposure Vans Excluding Hinds (in/a*2):
Location**
tutor 1, 2. or 3)
Otftslde Cloth 11
Average Quantlf I eat I on Llalt (ug/or2)
Hand Exposure Valuas (ug):
Ins Ida gloves: (R)
(L)
Beth
(I)
Both
Do Not L—
Outs Ida glovesfor no gloves): (R)
Average QuantlfIcrHon Llalt (ug):
Please Provide a Brief Description of the Incidental Exposure Situation
Ins do patch Is undar personal elothlM.
Inside patch Is undar protective clothln.
Inside patch Is undar both personal and protective clothing.
B-4
-------
Study Cob* Ne.r
(EPA Um Only)
Sampling Data
Tlae of Day Monitoring Took PI
F
MIXER/LOADER E»OSUR£ SURVEY FORM
(ONE FORM PER REPLICATE)
(mo allltary tlae):
To
CCMPQ0»p IQEMTIFI CAT I On 'Action of Pesticide: 1) Fungicide;
2) Herbicide; 3) Inaecticide; 4) Fu.lg.nt; 3) Plant Growth Regulator
•Formulation: Liquid (Ibai/gal) OR Solid (f al by Might)
1) Eaulslflabia Coneantrata
2) SuapansIon (aqeaous)
3) Mlcroancapsulatad
4) Solution
5) Liquid (undiluted; e.g., fwigant)
1) Matt* la-powder
2) Ory-flovable
3) Oast
4) Granule
•Packaging Type: 1) Bag
•Container Size
Marker
1.0.
•ork Cycle
Do Mat
Location of • Data-!"
*terj«ar Hork Cycl^
«J>.
2) Can
3) Bottle
4) Mater Soluble Packet
5 ITS Pftlfl. 'Location (State/ProvI nee and Country
•(chock one) ______ Indoor or Outdoor
MIXING/LOADING INFORMATION
•Mixing Procedures: 1) Open
If cloaad aystea Is ueed, specify
2) Closed, Mechanical Puap
and aodal
3) Cloaad, Gravity Feed
•Type of Diluent: 1) Mater 2) Oil, Vegetable
Final Mix Concentration (lb al/gal 01 leant)
t Tank Mixes Monitored
Tank/Hop pars Capacity (gal/lb)
Other Tank Addltlvlaa:
0)
1) Spreader/Sticker
2) Surfactant
Describe Mixing Procedures and Eqalpaent
3) Oil, Petroleua 4) Solvent
Total § Gal Ions Mixed
Total Aaoont of a I Mixed (lb)
3) Dye
4) Other Pesticide
B-5
-------
Morfcar
I.D.
ttork Cycla
Op Mat ntmik
Study Coda No.:_
(EPA U»a Only)
MIXER/LOADER EyQSURE SURVEY FORM - 2
Location
Morkar
1.0.
WMStiT
Horfc Cycl<
WEATHER DATA
% Ralatlva HumId Ity: High tw) ^
Toaparatura (Fahranhalt): High Lff1f
Wind SpMd: 1) 0-5 aph a) «-10 apfa 3) >io~^k ' ~~
Wind 0 tract Ion Ralatlva to Mixing: t) Faca to Kind 2) Back to vim 3) C rewind 4) Otfcar
ff 4, axplaln . • ¦ •
AIRBCRIg g»PSWE DATA
Tachi.1,. IM: 1) Flltar on ftaaplrator 2) Parsona! Air Pmp 3) Flxad Location
3, spacify location
*lr S_ll,g TIm (.1.) ,«wrHfl«n.. Ll.lt (q)
" CM.Ic.1 Col l«cr«d lag) Air
-------
Study Cod* No..^
(EPA U— Only)
MIXER/LOmew E)POSURE SURVEY FORM - 3
DERMAL nrWtt.1T tUTA
Stapling Tlaa (hrs)
***thod of Naasa-lng Oar«al Expo*or* Excluding Hands:
'> »*«teh Tachnlqat 2) Solvant RIum 3) Clothing 4) Othar
If 1. tacrlba patch eona+raetlon CI... 9.01., .|,ha-cal IuIom. ate.)
It 2, typa of aoivant
It 3, dascrlba clothing
>~ «, oascriba
•Krfhod of Measuring Hind Exposure: 1) Hand Rl 1
If 1, typa of *01 want
2) 61ov«
3) Othar
B-7
-------
Study Cod* No.:
(EPA Use Out v)
MIXER/LOADER ENSURE SURVEY FORM - t
Denial Exposure Vaoes Excluding Hinds (ug/ca?):
Location**
Ins Id* Clothing (wittr 1. 2. or 3)
Hud
Nock Front
Nock Back
Upper Arws
Shoul bars
Chast
Back
Forearms
Hips
Thighs
Shins
Cah
Ankli
Foot
(R)
(R)
CR>
(R>
(R)
(R)
(R)
(R)
(R) _
(R)
(L>
(L)
(L)
(L)
(L)
(L)
Average Quantification Llalt (ug/or2)
Hand Exposure Valuas (ug):
Ins Ida gloves: (R)
Outs Ida g lor as (or no glovas): (R)
Average Quantification Llalt (ug):
Outs Ida Clothing
(L)
(R)
(R)
CR)
(R)
(R)
(R)
(R)
(R)
(R)
(R)
(R)
(L)
(L>
(L)
CL)
(L)
(L)
(L>
(L)
(L)
JL)_
(L)
(L)
Both
Both
Location
Norkar
I J).
of • Data:
Mork Cycl«
1! peat lcIda uas observed, plaasa Ore la the rasponsa(s) that
dascrlba(s) the si tat Ion: i) Hand Cant act; 2) Tmea Contact; 3) otter Body Parts.
Pro* Ida a Br I of Description of the Incidental Exposure Situation
Pie
••Enter ana of the following nwrals corresponding to patch location:
1. Ins Ida patch Is undar personal clothing.
2. Ins Ida patch Is undar protective clothliq.
3. Inside patch Is undar both personal and protective clothing.
B-8
-------
Study
No.:
fiQilL
HI XER/LQADEW/APPLICATOR g»QSURg SIWVEY mw
(ONE FORM PER REPLICATE)
Sailing Oat*
T'** Monitoring Took Place (ma allltary tiaa):
To
CCHPOUM) IPgwriFtrAjiffl •Action of Pastlclda: 1) Funglcida;
2) Harblclda; 3) Intacticlda; 4) F»igant; 3) Plant Croat!) Raguiator
*Foraulatlon: Liquid (lb al/gal) OR Solid (Sal by Might)
') Eaulsiflabia Concantrala
2) Suapanalan (aquaeus)
3) Nlcroancapsalatad
^ ) Sol tft I OR
5) Liquid (ondlivtad; a.g., fwlgant)
t) Watt* le-powder
2) Dryfloaabla
3) Oast
4) 6 rang I a
•Packaging Typa: 1) Bag
*Containar Slz«
2) Can
3) Bottla
«om«r
I.D.
r
Sajlat^aixi_Biaak
Location of * Data*
Mortaar
1.0.
^aa inatrBeTTag"
Norfc Cycia
»r
4) Matar SoIobia Pack*?
SITg data 'Location (Stata/Prov inca and Country)
•(Oiack ona) Indoor or Outdoor *Crop
•Crop Halght (faat)
miximr/i fun.mr tirnamr.^
•Mixing Procaduras: 1) Opan
If cloaad systaa Is usad, spacify
. •Roa Spacing (faat)
2) Cloaad, Machanical Pta*
and aodal
3) Cloaad, Gravity Faad
•Typa of Dlluant: 1) Matar 2) Oil, Vagatabia
Final Mix Coneantration (Ibal/gal Oilaant)
I Tank Mlxas Menitorad
•Tank/Hoppar* Capacity (gal/lb)
Ottwr Tank Additlvlaa:
0)
t) Spraadar/Stlchar
2) Surfactant
3) Oil, Patrolana 4) Solvant
Total i 6aI Ions Mixod
Total Aaoont of al Mlxad (lb)
3) Dya
4) Olltar Paatieida
B-9
-------
Study Co*e Ns.:a
(EPA U— Only)
MlXER/LOADER/APPL ICATOR EXPOSURE SURVEY FORM . ,
Describe Mixing Procaduras and Cqulpeairt
nomer
¦or* Cyct« j
1.0.
Location of • Data: ™
¦orkar
1.0.
Work Cycle
Natter
iTKrions
APPLICATION [FORMATION
•Method of Application
l\ ^•r,fL* RotarY wln8 10) Shank Injection
?rOMn^.boai Ir"ctor 7) Lew Pressure Hand Hand 11) Fnaioatlon
l\ gr0Mn<||>0<» Truck 8) High Prwiar* Hand Hand 12) Solid Broadcast Spreader
^ <«ttached to truck) 13) Liquid Broadcast Spraadar
3) Aarlal - Fixad Ming 9) Backpack U) Othar(dMcrTba>*^
If 14, daaerlba
Rata: (lb al/acra) and (gal/acra) ___ Total lb al Appllad
Total § Acres Traatad Final Mix Ooneantratlon (lb al/gal Diluent)
Total t Gallons Sprayed § Tank Applications Monitor**
•Vehicle Maka and Modal:
*Cab Typa:
1) Opan Cab (Cockpit)
2) Closad Cab (Cockplt)/WI
Ground Spaad (aph)
3) Claaad Cab (Cockplt)/Vlndou Closad
Opan 4) Closad Cab (Cockplt)/Wlndo* Cloaed/Filterad Air
•AboIIcator Maka and Modal:
•Disc §
•Nozzle: Typa/Ma nu fact or a r
•Nozzle Prassora (pal)
•Swath Width (feat) ________
•Shank/Probe Typa: 1) Knives 2)
•Dapth of Injection/Incorporation (inches)
Describe Application Procad are
*t Nozzles
I/No.
Height Abore Ground (feet)
•Length of Bon (feet) _______
(3) Other
B-10
-------
Study Co* No.:a
ttPk IIm Only)
HIXER/LOAOER/APPLICATOR E»Q»E SURVEY FORM . i
•ark Cycle
Haaoar
¦ve Rla*
Location of • Data:
Worker
U>.
Worfc Cycle
and Low
WEATHER DAT*
S Relative Hwldlty: High __
Twperatnre (Fahrenheit): High
",nd 1) 0-5 aph 2) 6-10 aph 3, > 10
Hind Direction Relative to Nixing: 1) Fece to Wind 2) Back to VInd 3) Croasvlnd 4) Other
If 4, explain
and Low
Wind Direction Relative to Travel: 1) Parallel
If 3, explain
AJ
2) Perpendicular
2) Personal Air Piap
¦Eaasifif mm
*Techniq«a Used: 1) F!Iter on Respirator
If 3, specify location
Air Sampling Tlaa
-------
Stady Carta Mb.;
tEWA Uaa Only)
HI XER/LOAPER/APPl ICATOR E^OS
PEWHAL EXPOSURE DATA Danul Sampling Tlaa
-------
(gft IIm Only)
Ml XER/LOAPER/APPL ICATDR EXPOSURE SURVEY FORM - 5
Dannl Exposure Vaiaas Exc lading Hinds (ug/at2):
location**
ins Ids Clothing (antar 1. 2. er 3)
Hack Front
Nack Back
Uppar Anw
Shoul d*rs
Chaat
Back
Foraarm
Hips
Thighs
Shins
Calvas
Anklas
Faat
(R)
(R)
(R)
(R)
(R)
(R)
(R)
(R)
(R)
(R>
(R)
(L)
(L)
(L)
(L)
(L)
_CL>_
(I)
(L)
CL.)
(L)
(L)
Otftslda Cloth Inn
(R)
(R)
Both
Outslda glavas (or no giovas): (R)
Avaraga Quantification Llalt (ag):
(L)
Beth
If Incidantal contact with tha pasticIda aw obaarvad, pi
dascrlba(s) tha sltiatlon: i) * rrmtai I j 2)
clrela tha rasponsa(s) that
3) Ofehar Body
Plaaaa Prow Ida a Brlaf Oaacrlption of tha incidantal Expoaura Sltwtion
ponding to patch location:
**£ntar ona of tha following nMaraIs
1. Ins Ida patch Is andor parsonal clothing.
2. Ins Ida patch Is oidor protactlva clothing.
3. Ins Ida patch Is wdar both parsonal and protactlva clothlig,
B-13
-------
Study Cod* No.:^
(EPA Use Onlvl
Vomer
1.0.
*or% Cycim~
Do Mat L—y, Bijp,,
Sapling Date
Tlae of Day Hon I taring Took Pit
Fr
FLA6GER EffQSURE SURVEY FORM
(ONE FORM PER REPLICATE)
(UM ailltary tlae):
. To
J »
IDENTIFICATION *Actlon of Pesticide: 1) Fungicide;
2) Herbicide; 3) Insecticide; 4) Fualgent; 5) Plant Growth Ragulater
•Foraulatlon: Liquid (lb at/gal) « Solid (J ai by .eight)
1) Gaulslftable Concentrate
2) Suspension (aqueous)
3) Microencapsulated
4) Solution
3) Liquid (undiluted; e.g„ falgant)
1) MettAle-poeder
2) Dry-flovable
3) Dust
4) Granula
?ITS P*T* 'Location (State/Province and Country)
*Crop *Croa Height (feet)_
•Row Spacing (foot)
APPLICATION INFORMATION
•Matted of Application
1) Aarlal - Rotary Ming
•Airplane or Hal I copter Mafca and Model
of * Oata:'
•ork Cycle
Worker
1.0.
SeeTnatrucTToM"
2) Aerial - Fixed Ming
Rate: (lb al/acre) and (gal/acre) Total lb al Flagged
To+al * Acres Treated _____ Final Mix Concentration (lb al/gal Diluent)
Total § Gallons Sprayed ______ § Tank Applications Monitored ______
Ground Speed of Plana (aph)
•Describe Spray Equlpaent:
#0,"c * ———*§ Nozzles
B-14
-------
Study Cod* No.:
ItP* U— Only)
FIA08CT EXPOSURE SURVEY FORM - 4
Dcnul Exposure Vaees Excluding Hands (ug/ea2):
Location**
Inside Clothing
•enter
1.0.
¦ork^Cycle
Location
Worker
1.0.
¦SeTTnsl
of • Oat a:
Work Cycle
Nuafeer
Tactions
Head
Neck Front
Neck Back
Upper Arm
Shoulders
Chest
Back
Forearm
Hips
Thighs
Shins
Calves
Ankles
Feet
(R)
(R)
(R)
(R)
(R)
Average Quantification Halt (ug/o2)
Hand Exposure Values (ug):
Ins Id* gloves: (R>
(L)
Outs Ida glares (or no gloves): CR)
Average Quantification Llalt (19):
(L)
If Incidental contact with the pastleIda an observed, pi
dascrlba(s) the situation: x) land Contacts 2)
Please Pro* Id* a Brlaf Description of tha Incldantal
Both
Both
circle tha rasponsa(s) that
1 3) Otter Body Parts.
Situation
••Enter on* of tha toI loving nMorals correspond!ng to patch location:
1. Ins Ida patch Is under personal clothing.
2. Inside patch Is aider protect We clothing.
3. Inside patch Is under both personal and protective cloth(19,
B-17
-------
» ,
-------
Study Coda Na.:a
JSLJUS-SnLxLm
1.0.
•er^crci.
bo Hot lw»« Bla.*
Location of • Data:
FUGGER EyQSURE SURVEY FORM - 2
"iBortaar
1.0.
Herk Oycla
app< trfrifK itrnpyATinw
*Nozzla Prassara (pal)
Halght Above Ground (faat)
•Nozzla: Tvpa/Hanufactarar Modal/We.
•Swath Width (faat) _____________ •Laogth of Boa* (faat)
Oaacrlba Flagging Procadura: '
Mas Flaggar PoaItionad:
(chack ena) In Ti aafaiit Araa or On Parlaatar of Traataant Araa
HEATHER DATA
% Ralatlva Hwildlty: High
Taaparatura (Fahranhalt): High and Lea
Mind Spaad: 1) 0-5 aph 2) 6-10 aph 3) > 10 aph
Flaggar Pas It lor Ralatlva to Traataairt Araa:
1) Upwind 2) Downwind 3) Croaavlnd
«
B-15
f
-------
»*«*y Coda Mo.:
(EPA Um b»iti
ftornar
1.0.
*Qr* Cyci«"
Po Mat t—-1 gi.^.
FLAGGER EXPOSURE SWVEY FORM - ^
Mfamr rmuz pm T^,„iq» u^. t> Fnt#r R^>lr^OP
2) Personal Air Pap 3) Fixad Location
'* 3, spoolfy location
*'r Sampling TIm (ain)
A*>ant °* Chaalcal Col loetad (ug)
Hoa Rata (L/aln): initial
^^iow of * Data:
*yr I *?•* Cycla
aaa lintrqcf 1^-
Alr VoIum Col loetad CD
Final
D*fBal S"P«»aB Tlaa (hrs)
**Tto" °* a-M bpon tal.il,,
» NW.1-.N- 2> S)eun,K f}0ntr
•"rfl« of Nnxrl* mm i, hm
I* l» typa of soIwairt
2) Giovas 3) Othar
B-16
------- |