WAYNE PA F/6 PERCEPTUAL/PSYCHOMOTOR APPLIED … · ad-a093 981 applied psychological services inc wayne pa f/6 5/10 perceptual/psychomotor requirements basic to performance in 35

AD-A093 981 APPLIED PSYCHOLOGICAL SERVICES INC WAYNE PA F/6 5/10PERCEPTUAL/PSYCHOMOTOR REQUIREMENTS BASIC TO PERFORMANCE IN 35 -ETC(U)DEC 80 A I SIEGEL, P J FEDERMAN, E H WELSAND F33615-78-C-0032

UNCLASSIFIED AFHRL-TR-80-26 NL*4fllfllfllflfflfflflflflflflflflflflflflflflfl-mlllllllllllumuIIuuuuIuIuIu-IIIIIIIIIIII-mIIIIIIEIII--mEEIIEEIIEI

AFHRL-TR-80-26 L E E A

AIR FORCE o-J PErCEPUALps O .Ro o RolEQURE.MENTSI H BASIC] TO PERFORMANCE IN 35

H AIR FORCE SPECIALTIES

U By

Arthur I. Siegel'

M Philip J. FedermanAEugene H. Welsand

Applied Psych.logiia| Service,. Inc.AS -irn¢t" Center

Wase. Pennslvania 19087

NMANPOWER AND PERSONNEL DIVISION

Brooks Air Force Base, Texas 78235

RE December 1980

S

Final Report

0U

%ppi Pr .. f r plo I,"i r ..- l rimi. ., nliildD

C It

E A

S LABORATORY

AIR FORCE SYSTEMS COMMAND

BROOKS AIR FORCE BASE,TEXAS 78235[ 80.01 ,-

NOTIC:E

When U.S. Government drawings. speeifit-at ions. or other data are used for any purpose otherthan a definitely related G;overnnhent procurement operation, the Government thereby incursno responsibility nor any obligation whatsoever, and the fact that the G;ove~rnmient may haveformulated, furnished, or in any way supplied the said drawings. specifications, or other data

inot to be rgrebyipcaonor otews.as inay canr licensing the holder or any

other person or corporation. or conveyinog any rights or permission to manufacture. use, or sellany patented invention that may in any way be related thereto.

This final report was submitted by Applied psychological Services. Inc.. Wayne. PennsylIvania14M)87. under Contract F33015-78-C-0032. Project ILIR. with the Manpower and PersonnelD~ivision. Air Force Human Resources Laboratory, (AFSC). Brooks Air Force Base. Texas782:15. Dr. William E. Alley. Force Utiliz.ation Branch. was the Contract Monitor for theLaboratory.

This report has been reviewed by the Office of Public Affairs (PA) and is releasable it) theNational Technical Information Service (NTIS). At NTIS. it will be available to the generalpublic, including foreign nations.

This technical report has been reviewed and is ap)proved for publication.

NANC Y CtINN. Technical D~irectorManpower anti Personnel D~ivision

RONALDJ U. ThRR't. Colonfel. I'SAFComnlander

st'BjvC: TO EXPO)RT CONTROL LAWS

Tihi% document contains information for manufacturing or using munitionis of war. Export ofthe informiation contained herein, or release to foreign nationals within the United States.without first obtaining an export license, is a violation of the International Traffic in %rmsRegulations. Such v iolat ion is subject to a penalt- of tip to 2 years, iminprison men t anid a finte of

$IIMNI.N) under 22 U.S.C. 2778.

SECURITY C~L ASSIFIC ATION OF THIS PAGE (54'., Dill. F.,1.,.dJ

igREPORT DOCUMENTATION PAGE BSOFCMLTN -,~2 GOV2T ACESO NO 3 RPIFN T' C AT ALO(; NOMBF;7I

S4 TITLE 5~ ACTSO O YP-E OF REPO9RT & PFRO(C) ,,j l

T Atrj. aF -)IT9RACQ OR OiSANI "M,2

I) P~OPMlNG nltAFtZATION NAME AN[) ACORI' - T'OA -E tYI N' -!ARE A d WC Re~ jr, I t~i.

I I CONTROLLING Ol' FICE N AME AND ADDRESIS 12 "P'OR"I GAIki

III9,,k- ki9 I.9on, Rik. ', .09 .8 iU BRGE P ,.

14 MONTO(RING AItNCY NAME AOI ,JIo,'.,-~ 9. Il. I5 i.E J R, T , .L ASS ~I',o.

llwok I 'kit IF't

16 OISTRiI8UTIO-N STATEMENT re tIh/ig,I

79 SUPPL.EMFN IAR-1 N7)?IL4,

j19i I 911 11 I w91411101Ir l UI 91.49*. Ik.X pila11%(

9., I.- IIII .~,9911I 9.

NI II. I .1, , .Igt I91,19 I.49 fill, .Rd 111 .. 1 I 4 lk,1 hit 9I. .l 91 9191I I 9I

.99j91II.91.1.IdSIIl , ma .. I- ilia III It99 9 9411, 1 , I ;dl,,44,4., I4M .I.rIr4.I ......

-. .. ....9 ..

DO 999791473 1. 4. '

1 9lI A ',,.Il N -1iS "'Al IN..

SECURITY CLASSIFICATION, OF THIS PAGE(Wlsmeui 03.10 n~r

field., are-. ill part. u ioi tie. X factor anal tsi of tilt- dala ioid icaled that tilt-Jeip ak/~seitltitrai tIx oiioiiil call lie, 4de-sribed hi% thIree facior-: %s ial. audition. and ma inual fators.

The cituiiii of do- slt% p~ oinut to: (a) tiaIII( iuWalu Of tile lamoinoi G I riing tlt-i kercepittal/1 i-~thiiiitttir requ~iremet.it ofi %ir hiVom career fit'li. 1,) tdeo ij~t detiiIt~ ei 1 Iii dfrftr

iimvefigatioo., of 1iii, sort. (c) thet partiakll luilit i rttia/elhiiIo requireiienis for arius Xir Fortecarver fiekdk. wni (dk) tlt- adeqiiati of tile piresent effort as a biasis for lic rcepituah/rpstiicioottr test dte elopoiuwii furlisit, allai aid iii tarver seetion and classificalioI.

I tnla'-ielli

SECURITY CLASSIFICATION OF THIS PAOE(WIhen D610 Ente,.d)

PREFACE

This research was conducted under Work Unit ILIR0042 (LabDirector's Fund), Development of Methodologies for Assessing Percep-tual/Psychomotor Ability Requirements in Air Force Specialties. Ap-preciation is expressed to Mrs. Mary Spencer (AFHRL/MP) for her as-sistance in assembling the task inventory data and to Mr. Henry Clark(AFHRL/TS) for his outstanding efforts during the sample identificationphases.

. TA,.

AV- i

Dizt i L

Avia l~ ',,,

-7=7.-17 L-ups=-,

TA. I I\~NO1 F:~\ I,' \T 1:, NI

n r .) m ci ..

I~~ .~SII~I .- . . . . . . . . . .

Pc k11 alW Psy hil l .u I .r~ .Iz . .~ . . . .

6\~ i~ 'B . . 0 Ui . h .t . .~ . A . . . . . . . . .

d. .L . .u. . .I. . . .I.\.I. .I... .

Pi3\I1O ~ i. .i .bI~ . .I . . . . ... . .

lhtoiet Isue I I

ke I sittiv Scain A p t I I II,

1lpesof eeontinua. .... . . 2

B eatig e Scale oAntie . ..3

M naiae When1011(,It S k' I tI I o IB tu~t l S i 0

111.rta Isue .''11 .)E .L ~ .N A. P. . l . .. ..........

Typeo s of Jo nha ...y ... ... .. 2

lic~atin One Sc le to A ille ........ 3

TABLE OF CONTENTS (cont.)

Page

Direct Observation by Analysis ............... ... 39Interview Methods .............................. 40Questionnaires and Task Lists ...... ............ 41Daily Diary ........ ...................... ... 42Critical Incident ...... .................... ... 42Comparison of Various Approaches ........... ... 43

Task List Method .......... ..................... 44

Discussion of Task List Method ..... ........... 44

Pretest Sample and Task Sample ................ .... 46

Pretest Sample ....... .................... ... 46Task Sample ....... ..................... .... 47

Development of Forms ......... .................. 48

Perceptual/Psychomotor Abilities ............. ... 48Perceptual/ Psychomotor Ability Requirements

Questionnaires ...... .................... ... 55Pretest Administration ..... ................ ... 56Administrator's Training ....... ............... 56

Administrator Instructions ........ ................ 57

Personal Interview ...... ..................... .... 59

Interviewer Instructions .... ............... .... 60

Interviewer Instructions ...... .................. ... 60

Interview Sample ......... ................... 61

Results--Quantitative ......... ................... 61

Discrimination- -Taxonomic Classes- - FireProtection Career Field ................... ... 62

Discrimination- -Taxonomic Classes--MunitionsMaintenance Career Field ...... .............. 65

iv

-- nor

TABLE OFCONTENTS (cont.)

Discrimination- -Tasks -- Fire Protection. .. ...... 68Discrimination- -Tasks- -Munitions Maintenance 68Discussion and Summary of Discrimination

Analyses.. ........... ........... 71Amount and Performance Quality Variability

Comparisons. .. ......... ........... 71Supervisor and S-ibordinate Comparisons. .. ...... 74Test-Retest Reliability .. .. ........ ....... 76iterrater Agreement .. ...... ........... 78Intercorrelation Among Abilities. ... ......... 81Categorical and Magnitude Estimation Scale

Com-parisons. .. ......... ........... 84

Properties of Scales. .. ....... .. .......... 90

Combined Profiles. .. ....... ........... 90

Interview Findings. .. ........... ......... 93

Discussion of Pretest Results .. ....... ........ 98

IV. PERCEPTUAL! PSYCHOMOTOR REQUIREMENTS O1F35 AFSCS. .. ....... ........... ...... 101

Changes in Forms and Procedures .. ..... .... 101

Sample. .. ........... .......... .... 101

Career Pields .. ..... ........... .... 101Air Forc e Bases .. ........... ........ 102Hespondent Sample. .... ........... ... 105Task Sample. .. ....... ........... .. 105

Data Collec-tion Instruments. .. .......... .... 100

Content of Data Collection Uorms. .. ..... .... 106Demographic Information. .. ......... .... 1 06

Procedures. .. ....... ........... ..... 106

Preparatory Step. .... .......... ..... 106Ad ministrator Instruc-tions .. ....... ....... 108

Demographic Description of Sample. .. ....... ... 108

v

FA I 'I 0 1,

1S Oi r v I w I . . .I. I. .t . . . . . .

S. . . . . . ...... . . ......... ..," u .. i . llt. . . ),. l }, .l. . . . . . . . . . . . . . . . . . . . .

.~ .I .i .. . .I..I ..

' m1 1 ' 10 il 1 1 It

.(. . . . . . . .

\) k I .I...................................

la, v. .I ' .I . . ., .' . L it. . . . i.. . . . . . . . . . ..

Cl : I ' V. . .. 1 !.' U . . . . . . .I

"!tl~~~~l:"~ .c ".o\, .\~ l ! '. , t l . .)ll . . . . . . . . . ..

" -- I* i ,' ' ., -tl .. J itv ]l . ih

t'lt ,

I't I I <, I 1 , f -

" .. .l i . . . . . . . . . :

t 11 t I i111F cioI

A N* I

itir

>-'i . ................. ......

N ilei )l' i'~ .l \i ,, A ~ t', hih l- lol~. '! "

'N Nt I\ '& ; , ) I I I- I I ,' t I,, u in iik Q A

. . . . . . . . . . . . . . . . . . . . ,

\ ' ' I s iI) ' t- I t'I,'. ,' I -, I~ t -, -!, ' t t ," li,"

ll t 'l I Xl i I *t . I I,

\'., l~tl ill+ ,l; lll i , ! ), 1 ' l ', , ;~ l > , l l l l l \ / l l , < l

TabhI iLL

2 -1 Mts-sificatoi tit 11c.iivi.t A.. wV H

ifivaui Sslyt-m 0iblL VtL..

2-2 SummI~ary ofli 1hruro'v Fax. *nwn' ..

P'erceptual ilatt cl by L';.............

2-4 R~otat ed ld& II'> ) hltilk~l I: k.'11

andi the ai' iabhi'! P't' L I I)' iI I 11.' A hk ,

rAv yordm in *o k~. w it .~ .in . .il .. .1

2 G t'ei' ii f n Pevc '1Q''l hn>d ") mAll

3i-2 ruask Sampjtmgt Pu't),a : w

I )CV atl H '~1 01. I 1 i l ,'* I tm i. i ' . .

3 -4 !Xihlit W~tans ai Stawh v Mvt i n m: 'Iil

M ItItilt 101 MV a [tit kilail. (', k'(' . . . .

III( 1 ilv . . . . . . . .

M\ l ilt ('1,11k Cal efl' t'litr .I . .

11 At 'V~tp k ~ I ' ~ I. L' I I .. . . . . . .

I l I St I t i t I d I l i t> ( ' IN F I I A l t I I ' , ., 1

I I

LIST OF TABLES (cont.)

Fable Page

3-11 Percentage Agreement Among Raters for thePerceptual! Psychomotor Abilities on theCategorical Scale Judgments .... ............... .... 79

3-12 Percentage Agreement Among Raters for thePerceptual! Psychomotor Abilities on theMagnitude Estimation Scale Judgments .......... .... 80

3-13 Frequency Distribution of IntercorrelationsAmong Taxonomic Classes for Amount .. ......... ... 82

3-14 Frequency Distribution of IntercorrelationsAmong Taxonomic Classes for PerformanceQuality Variability ....... .................... .... 83

3-15 Product Moment Correlation Between Categoricaland Magnitude Estimation Scaling Approaches(N = 13 Abilities) ....... ..................... ... 85

4-1 Sample of Air Force Career Fields by AptitudeScore Minimums ....... ..................... ... 103

4-2 Air Base Sample ........ ...................... ... 104

4-3 Frequency of Supervisors in the Final Sample in35 AFSCs in Four Aptitude Areas with Medianand Mode of Years in their Specialty (AFSC)and Years in the Air Force (AF) .... ............. ... 110

4-4 Frequency of Subordinates in the Final Sample in35 AFSCs in Four Aptitude Areas with Medianand Mode of Years in their Specialty (AFSC) andYears in the Air Force (AF) ..... ............... ... 112

4-5 Means and SDs of Judgments Concerning Amountof 13 Perceptual/ Psychomotor Abilities Combinedfor Supervisors and Subordinates in 8 AFSCs andin the Mechanical Aptitude Area ...... ............. 114

4-6 Means and SDs of Judgments Concerning Amountof 1 3 Perceptual/ Psychomotor Abilities Combinedfor Supervisors and Subordinates in 9 AFSCs andin the Administrative Aptitude Area .... .......... ... 115

4-7 Means and SDs of Judgments Concerning Amountof 13 Perceptual/ Psychomotor Abilities Combinedfor Supervisors and Subordinates in 9 AFSCs and

in the General Aptitude Area .... ............... ... 116

viii


Table Page

4-8 Means and SDs of Judgments Concerning Amountof 13 Perceptual/ Psychomotor Abilities Combinedfor Supervisors and Subordinates in 9 AFSCs andin the Electronics Aptitude Area. .. ... ..... ... 117

4-9 "High" and "Low" Amount of 13 Perceptual/ Psycho-motor Abilities in 29 AFSCs by Aptitude Areas . . . . . 120

4-10 Means and SDs of Judgments Concerning PerformanceQuality Variabilities as a Function of 13 Perceptual/Psychomotor Abilities Combined for Supervisors andSubordinates in 8 AFSCs and in the Mechanical Apti-tude Area. .. ... ...... ...... ..... ... 124

4-11 Means and SDs of Judgments Concerning PerformanceQuality Variabilities as a Function of 13 Perceptual/Psychomotor Abilities Combined for Supervisors andSubordinates in 9 AFSCs and in the AdministrativeAptitude Area...... ..... . . . .. .. .. .. .. .. .. 125

4-12 Means and SDs of Judgments Concerning PerformanceQuality Variabilities as a Function of 13 Perceptual/Psychomotor Abilities Combined for Supervisors andSubordinates in 9 AFSCs and in the General AptitudeAre a. .. ..... ...... ...... ..... ... 126

4-13 Means and SDs of Judgments Concerning PerformanceQuality Variabilities as a Function of 13 Perceptual/Psychomotor Abilities Combined for Supervisors andSubordinates in 9 AFSCs and in the Electronics Apti-tude Area. .. .. ...... ...... ..... ..... 127

4-14 "High" and "Low" Performance Quality Variability asa Function of Perceptual/ Psychomotor Abilities in13 AFSCs by Aptitude Area .. ... ...... ...... 129

4-15 Frequency of Means of Judgments of Tasks Relativeto Amount of 13 Perceptual/ Psychomotor AbilitiesOne-Half SD Unit Above and Below Relevant Meansin 8 AFSCs and Mechanical Aptitude Area. .. .. ..... 133

4-16 Frequency of Means of Judgments of Tasks Relativeto Amount of 13 Perceptual/ Psychomotor AbilitiesOne-Half SD Unit Above and Below Relevant Meansin 9 AFSCs and Administrative Aptitude Area .. .. ... 134

ix

yd

Aiim (d 1 ' P.l1ki/ cyliotnotor Abli liiI, HAI' 11.1 SIl it\l itd Bt'Iow Relevmirt Nleir;

1 , 'r(t'jttw ty ofJ]n' of. lA inf idst of Fais I:; Rt ii j \

i Au 0)11111 ()!I ]A I '' 'i t ti, 'v l n t AL il it

iii I I iif ;DI) lil Altttv and Plelow 16,ievrt Nlettlin 91 AF1SC s anti 1 Ii.' eiics Aptitude Aruc' . . . i..t

-9 r'w.H ['i.sks \Vhi. h Worte llitliest on E~ach I'ei-eptul/

4 -2 NI' in )f Ji id itt ti it ,fit isks-- Relative t I1ifti'tn n 0

k.Imi . i v tr i lil lily 1- ji I nic tion of 1 3 Pt rce pt ut]l

P'. ti~ ,,, Il hit je oine'-11l af SDI U nit Altovi' iulI

1'. l'-% Niii i 8 \F'SCs and Mtc,wicA Ai p--

A ~ ~ ~ ~ ~ ) I Ii IIi'isHtsks; Relaitive to P'erfowrmall'IQi111l i Ln Ii ihjlV "(s- j Fulintion of 1:3 peluteptual/

P:c tii )It ihi-i ( ne-Half SD) Unit Above andI

Bel''. 1 'V imi Nlt ttis in 9 AFSCs and Adinis,,trative

A Nt. tnt of I)1 t tn 1 ':sks Relative to Pe rformnanei t' v 1 Itib i \ iS Uu l c t ion of 1.3 Pe rc 1 p)t ra I

li~li~t~ h ihiwin One -Half SD Unit Above anld'.1ev W~x 'i mit,itt ill 9 ArFSCs and Genieral Apti -

.- ,"I Ninm: )f J 1ete 'sks lielalijve to Pe(rforlanteIi y N I i . -a 1iiition of 1:3 1 I'Vepf1tial/

1' ty~h~ttitt tt NH One -Half SI) Unit Ahove arid

B'VwRelvan NI.'.Int 9 Al S(.s and 1-leetronlikts Ap-

I 2 I \tt'~w'w ltlite.'; "etfween M'ean,- of Jiudgments 01.\'p1nt i -1 lO'vitiatt ality Variaibility . . . ..... I

1,' ct11ii n~ ( It ti roit I,' it' \ltrIit y M/eans for A mount

IN II i i .-\...............................1.....................II I

N


Table Page

4-27 Correlation Between Supervisor and SubordinateRatings for Amount (A) and Performance QualityVariability (PQV) Judgments .. .. .. .. .. .. .. 161

4-28 Percentage of D Scores in Each AFSC Within 10

Percent of the D Range (2. 52) or Less .............. 163

4-29 Intercorrelation Matrix of Amount Ratings AcrossAFSCs and Tasks for Each Ability .............. ... 164

4-30 ANOVA Model for Reliability Determination .... ...... 167

4-31 Reliability of Subordinate and Supervisory Ratingsfor a Variety of Sample Sizes and AFSCs ......... ... 169

4-32 Formulas Employed for Estimating Strength ofAssociation (w z ) Between Independent and De-pendent Variables ....... ..................... 171

4-33 Strength of Association (w 2 ) Between Independentand Dependent Variables for Subordinate Raters .. 172

4-34 Strength of Association (w 2 ) Between Independentand Dependent Variables for Supervisory Raters . ... 173

xi

LIST OF FIGURES

Figure

3-1 Correspondence between two scaling approachesfor the [ire Protection specialty on amount ....... 86

3-2 Correspondence between two scaling approachesfor the Fire Protection specialty on perform-ance quality variability. ............... ........ 87

3-3 Correspondence between two scaling approachesfor Munitions Maintenance specialty on amount . . . 88

3-4 Correspondence between two scaling approachesfor Munitions Maintenance specialty on perform-ance quality variability. ............... ........ 89

3-5 Means and standard deviations of taxonomic classes'or 'ire Protection career field ... ........... ... 91

3-6 Means and standard deviations of taxonomic classesfor Munitions Maintenance career . ............ ... 92

3-7 Involvement of perceptual/psychomotor abilitiesin ['ire Protection career field ............... ... 94

3-8 Involvement of per'cptual/psychomotor abilitiesin Munitions Maintenance career field .......... ... 95

4-1 Mean and + .5 standard deviation for each abilityfor judgment of amount ..... ................ ... 145

4-2 Mean and +. 5 standard deviation for each abilityfor judgment of quality variability ............. ... 151

xii

LIST OF EXHIBITS

Exhibit Page

4-1 Scales used for parts 1 and 2 of the data collectionfor ms ....... ........................... ... 107

A-1 Sample response page from the data collection forms . . 196

xiii

I. INTRODUCTION

A major problem in advanced, technologically oriented organiza-tions, such as the Air Force, is the determination of the ability require-ments for job performance. Job descriptions, in terms of ability re-quirements, support a variety of purposes, such as establishing job en-try requirements, classification and assignment, performance evalua-tion, training requirements derivation, and career development.

This report presents the methods, procedures, and results of astudy which sought

1. To develop a perceptual/ psychomotor taxonomy ofabilities applicable to Air Force career fields.

2. To apply the taxonomy to a variety of Air Forcespecialties in order to describe the perceptual/psychomotor abilities inherent to the job perform-ance of personnel in these specialties.

While the Air Force routinely collects information about jobcharacteristics through its Comprehensive Occupational Data AnalysisProgram (CODAP) system (Christal, 1974), this information has notdealt with perceptual/ psychomotor requirements. Moreover, althoughthere has been substantial progress toward predicting success in AirForce technical training schools and (to a lesser extent) job performanceon the basis of academic/cognitive skills, there has been little, if any,emphasis on prediction based on perceptual/psychomotor abilities. It isquite possible that inclusion of perceptual/psychomotor ability consider-ations in such predictions would substantially enhance predictive power.This holds, because, on the surface, it seems that perceptual/psycho-motor ability is basic to a large part of the performance of Air Forceenlisted personnel.

The objective of the present study was to describe the perceptual/psychomotor requirements of 35 Air Force specialties. However, beforesuch requirements could be established, some basic needs had to be met,including:

1. A perceptual/pscyhomotor taxonomy that is defen-sible, comprehensive, scalable, and applicable tothe work of Air Force career, fields.

2. A method for economically collecting thu informa-

tion dictated by the taxonomy.

I

3. An information collection method that met accept-able measurement standards and was reasonablyfree from error.

Subsequent sections of this report describe the steps taken tomeet each of these needs. Then, the methods and results of a surveyto establish the perceptual/psychomotor requirements of 35 Air Forcespecialties are described. These data can now provide a basis for fu-ture programs which exploit perceptual/psychomotor performance re-quirements in the Air Force.

2

II. TAXONOMY DEVELOPMENT

A taxonomy is a "classification of data according to their natur-al relationships or the principle governing such classification" (English& English, 1958) or "a way of simplifying a complicated universe of in-dividual events and objects according to some useful way of identifyingthe way in which groups of individuals (or observations) have things incommon or differ" (Miller, 1967). In short, a taxonomy is a classifica-tion system. The desirable characteristics in an Air Force oriented be-havioral taxonomy, as extrapolated from Miller (1967) and Fleishman(1975), are

1. Compatibility--the scheme should be fully com-patible with the Air Force task structure.

2. Understandability- -the scheme must be readilyapparent and comprehensible to Air Force users.

3. Objectivity- -the standards for evaluation mustbe free from bias.

4. Scalability--the technique should allow for the as-signment of a magnitude value (a number) to thetasks of a job relative to each class in the scheme.

5. Practicality- -the scheme should be relatively sim-ple to apply and interpret and should not place un-due time requirements on operational personnel.

6. Validity--the scheme should be based on accepta-ble constructs relevant to Air Force job content,and seem reasonable to the Air Force users.

7. Reliability--the scheme should be amenable to psy-chometrically reliable data acquisition methods.

8. Comprehensive, generality, and flexibility- -thescheme should be applicable to the full range oftasks involved in Air Force career fields.

9. Cost effective--the taxonomy should have charac-teristics that permit it to be embedded within ascheme that is relatively inexpensive to employand the taxonomy should be purposeful in estab-lishing an appropriate job-personnel interface.

10. Unidimensionality--each skill within the schemeshould be unique.

:3

Perceptual/Psychomotor Taxonomies

A number of rather general taxonomies have been developedover the years (Ragsdale, 1950; Fitts, 1962; Simpson, 1972; McCormick,Jeanneret, & Mecham, 1969; Kibler, Barker, &Miles, 1970; Greenstein,1976). These systems attempted to establish broad categories and classesinto which behaviors could be classified. They do not deal with percep-tual/psychomotor abilities per se and, therefore, are not reviewed here-in.

Major Perceptual Motor Ability Taxonomies Prior to 1968

Prior to 1968, the most important taxonomies developed utilizingperceptual/psychomotor descriptors were those of Berliner (Rabideau,1966), and Fleishman (1967).

Berliner's Taxonomy

The first formal task classification system utilizing perceptual/psychomotor descriptors is the Berliner three-tier system (see Rabideau,1964). The Berliner system, shown in "able2-1, classifies tasks in termsof intervening human processes or functions as related to general workactivities and specific behaviors or tasks. The Berliner taxonomy wasdeveloped to organize and define the measurement of task performance,and provides a useful scheme for analyzing man-machine systems.

The perceptual motor components in the Berliner system corr-e-spond to the processes identified as perceptual and motor. Using suchdescriptors, the system describes man-machine performance in termsof human behaviors, and outlines the human processes required to per-form the tasks. No detailed definitions of the processes are given. Themajor emphasis is on the human activities. Accordingly, the Berlinertaxonomy can best be considered as a behavioral description rather thana behavioral requirements approach to task classification.

Several other authors have followed Berliner in developing simi-lar types of taxonomies, for example, Alluisi (1967) and Chambers (1973).As with Berliner, these two taxonomies utilize perceptual/psychomotorfunctions, but emphasize the human activities or tasks. Such taxonomicsystems suggest the perceptual/psychomotor abilities required to performtasks. However, the taxonomies are too broad or vague to identify theamount or type of perceptual/psychomotor ability required. Taken togeth-er, the three systems appear to be descriptive, nonrigorous, qualitative,and general types of taxonomies. They were subjectively developed anddepend heavily on verbal descriptions, with somewhat overlapping func-tions and behaviors.

4

Prcw! .' A I I

11111." for lmte c t i ris P

r "~ iVi112 111" -11 :-w've, read,

kint ift [Iup _. ts, 1. 1 )iu; rim iriati.

0 it 1

rc Ii ('i IV '' .O)

I I n IcC '-1'

Motor~ I col

LI r) U Oi'VP ; P

Fleishman's Taxonomy

Fleishman and his coworkers (1966, 1975) identified a number ofmajor perceptual/psychomotor abilities as accounting for performance inover 200 types of tasks:

I. Control Precision--muscular adjustments of the largemuscle groups, e.g., arm-hand and leg movements.The ability is very important when quick and preciseadjustments are required.

2. Multilimb Coordination- -involves the use of arm-handand leg movements. The ability is involved in per-ceptual-motor tasks where the ability to coordinate

the movements of the limbs simultaneously are re-quired.

3. Response Orientation--the ability to move quickly andcorrectly in response to a stimulus.

4. Reaction Time--the time elapsed between the appear-ance of a stimulus and the response.

5. Speed of Arm Movement--the time required to execute

an arm movement, where accuracy is not required. !6. Rate Control (Timing)--timing motor adjustments in

response to changes in moving targets or objects.

Compensatory aid pursuit movements are involved inthis ability, as well as the ability to respond to changesin rate.

7. Manual Dexterity--the ability to perform tasks involvingthe manipulation of large objects. Skillful arm-hand

movements under speeded conditions are also involved.

8. Finger Dexterity--the ability to manipulate small ob-

jects with the fingers.

9. Arm-Htand Steadiness- -the ability to make accuratemovements involving the arm and hand. Only stead-ines, is an important factor; strength and speed arenot involved.

10. Wrist-Finger Speed (Tapping)--the ability to makerapid wrist flexing and finger tapping movements.

Extent Flexibility - -forward, backward, and lateralstretching or flexing of trunk and/ or back muscles.

12. Dynamic Flexibilit2 -- the ability to repeat rapid mus-cle flexing movements. The ability is critical in therecovery from strain or distortion.

13. Static Strength--the force exerted against objects fora brief period of time. The ability is not involved insupporting one's own body weight.

14. Dynamic Strength- -muscular resistance to fatigue.The ability involves muscular endurance when con-tinuous muscular force is required over i consid-erable period of time.

15. Trunk Strength--the ability involves the trunk andabdominal muscle resistance to fatigue, when useof these muscles is required over a considerableperiod of time.

16. Explosive Strength- -the ability to engage in activitiesrequiring short bursts of energy and strength. Con-tinuous stress or strain, through repeated exertion ofmuscles, is not involved.

17. Gross Body Coordination- -the ability to coordinate theactions of different parts of the body while simultane-ously executing gross body movements.

18. Gross Body Equilibruim- -the ability to maintain bodybalance while opposing forces are acting on the body.

19. Stamina (Cardiovascular Endurance)--the ability toengage in physical activity and expend effort overlong periods of time.

Fleishman and his colleagues formulated the perceptual/ psycho-motor abilities and their definitions through a series of interrelated ex-perimental, correlational, and factor analytic studies. He provided (1967)an example of his experimental -factor analytic paradigm by describingthe process in the development of the Rate Control ability factor. Hefound that in early studies this factor was common to compensatory aswell as pursuit tasks. To test the factors' generality, tasks were devel-oped to emphasize rate control. The factor was found to extend to suchtasks. In later studies the attempt was made to discover if emphasis onthis ability is in judging the rate of the stimulus as distinguished fromability to respond at the appropriate rate. A task was developed to in-vestigate this conjecture. Performance on this task did not correlate

7

I k s~ in0 1( li w l I I v I i i \ I ti r' K' bt k \

I I A li 'Itkl x ir lpo~ i~i I i t , ,I"( , i t I )\ Ii I i

it kit LIs wa foun 1[i t11 ,i ii~ k i ii k i k'

I ito t t to 11tkl ud ii I ICttt i llk-- htl i' Iti t 'ii: II

* ~ ~ ~ ~ ~ -:'1 ri. I.: i Iql \i't ' Il 1 iO ' . i

t- u1

t II t' oI

it x !'l x 97 1, lI Ii: .I-i'

ra

I .1)1'' 2 :1

. -, ,,. , ,..,.-,,i' I

ateg)r\

AK

~ Move 'P

I oP:p ..I * . . $

4dm0

- eba.

Streu

I' Ic x.

:1 , .. .

(I

-e

Table 2-2 (Cont.)

5. Skilled Movements a. Simple Performing tasks of Activities Wriich build on "Bas-intricate and involved ic Fundamental Movements

b. Compound movement based on (Category 2) as in athletics,basic movement pat- jancing, and painting

c. Complex terns

6. Nondiscursive a. Expressive Communicating .nrough Posture, body gestures, gri-Communication movement body movements and maces, sign language, and

facial expressions danieb. Interpretive

movement

a Adapted from Harrow (1972).

10

6. Figure-Ground Discrimination--the ability to se-lect the dominant figure from the surroundingbackground.

7. Perceptual Constancy--the ability to recognizefamiliar symbols when presented in a novel man-ner or size.

8. Auditory Acuity--the ability to discriminate be-tween sounds of pitch and intensity differences.

9. Auditory Tracking--the ability to locate soundsand track their direction and movements.

10. Auditory Memory--the ability to recognize andreproduce (e.g., verbally) past auditory experi-ences.

Though restricted more to the perceptual side of perceptual/[chomotor abilities, Harrow's 10 abilities can be included in a more cprehensive list of perceptual/ psychomotor abilities. For example, tican be included with Fleishman's 19 abilities, which emphasize the mtor end of the perceptual/ psychomotor ability dimension.

Hunter's Taxonomy

Hunter (1975) defined 11 ability factors through factor analysisof the results of seven perceptual/psychomotor apparatus tests and 21paper-and-pencil tests. A major purpose was to compare the factorialstructure of apparatus tests with the factorial structure of paper-and-pencil tests. The results showed that the two test batteries shared littlecommon variance. Six ability factors were specific to the perceptual/psychomotor apparatus battery, four factors were specific to the paper-and-pencil measures, and one factor was common to both batteries.First, seven factors were extracted and rotated from the perceptual/psy-chomotor apparatus battery analysis: (a) Visual Tracking, (b) AuditoryTracking, (c) Figural Memory, (d) Position Memory, (e) Motor Speed,(f) Associate Speed, and (g) Perceptual Speed. The factors were definedby those apparatus tests (labeled 'rest 1 to 'rest 7) which loaded highly(i.e., >. + . 30) on the respective factors (Table 2-3).

When Hunter factor analyzed the combined perceptual/psychomo-tor apparatus and the paper-and-pencil batteries, 11 rotated factorswere obtained (rable 2-4): (a) Verbal, (b) Spatial Relations, (c) VisualTracking, (d) Figural Memory, (e) Auditory Tracking, (f) Mechanical,(g) Associative Speed, (h) Motor Speed, (i) ManualDexterity 1, (j) ManualDexterity 11, and (k) Perceptual Speed.

11

TI I I'

Unde ~ ~ ~ r I It<,'At r

to-Ln ftro 9.f.

T,c P~-riot gjn, r

Tc'st 7 P',tur za I,-L de r P Ividdd Att- is-,

ri on- I itte FrrotNM..rku e I

' -, r ,F (i Pe ~ t .I

Ltde 'F.'rdo (Dtlei;,'d0 A

r('s rt De I , 4

l' irr 2

t 3,t.rFor-manc,rre iuteC~Ans- -

Re-;ponsc I )xq

mIemory- Corr* Se teI 0is

?erceo r on Timne uo Pv r c i-p t,.,, L v ri

J Ictotr V (A-4iiocijitIieF:rU ~sSpseed) S~d

Te'st 4, Associative Test 3, V~ LcLealrning PVsrt 1 -69 St70S-P'r eirt't toss '

Tt 1, Kines !.te Ic, 2, Pr r--.-Prssjl ''

Meory- Response Time .hS Percept~ion Tim~e.i.est 4, Associative

Learninsg Part 2 -5rest 6, Concept Idenri--fication-Correct Answers -4

Tesr: 2. Perceptual Speed-?erceptton Time . ______

Taken from litnter (1975).

I if ( i)l d r tir, i

1'5\clIOnlutuP/l'cli)tutI Ittef\ Ik. I

i:ic tor Fa C t o r

Vari ab I ,.od in\ar ui bl L L.oa d ir.

Factor I (Ve rbal) Factor Ii (SpatialRelations)

Word Knowl dg, -. 81 Hidden !.iguresVerbal Analogies -. 76 Pattern DetailWord ;rouping . 73 Figure Analo ies . ,Il'etter Sets .56 Rotated Blocks . T,S:alu Readi Z- ) 5 Electrical MazeI lectrical Inftrmit ion -. '5 Block (lnuntin, .

I ,Ole Reading -. 40 Point Distmnct . 3kl,. hanica I Principlt,,; -. ,_0 Letter S ,tsFicurc Analoies -. 0 Scale R.ading .

;st 6, Concept Idintificn- Mechanical Principlts .

t ion-Corr,,ct Ansvrs -.

Block Count in-,

.,lctor I1 (Visal Factor IV igur;lTrack inv) Memo r v)

Test 7, Pertrman, " lnder Test 5, Memorv Part Itt'-id .t A L - i , -~i , (Immc~dialtt) . 1,4

Error, Minute, 2 Test 5, P Ieorv I'Lart 2Test ', PerIormnc,-, (1 laved) I '

Und,.r !ividc.] Attention- Test 1 , KinestheticL.int, Lrror, Minut.' i .91 Memo ryv-Correct Answers . it

Test 7, Peri ormc.:T. Test 3, Perl ormlnceUnder I)ivided Attention- Under Stress-CorrectLine Error, Minute I .6 Answers .50

Test /, Performance Discrimination-React ion .49Under Divided Attention- Answer Sheet Marking,Line Error, Minute 4 .88 Rights .49

Table Reading .49

Test 2, Perceptoal

Speed-Response

Time -.46Test 2, Perceptual

Speed-Correct

Answers .44Block Counting .37

Point Distance .36Test 2, Perceptual Speed

Perception Time -.33Letter Sets . 1

13

Trable 2-4 (cont.)

Factor V (AuditoryTracking Factor VI (Mechanical)

Test 7, Performance Under Tools -.75Divided Attention-Tone Tool Functions -.74Error, Minute 2 .81 Electrical Information -.62

Test 7, Performance Under Mechanical Principles -.55Divided Attention-Tone Electrical Maze -.31Error, Minute 3 .81

Test 7, Performance UnderDivided Attention-ToneError, Minute 1 .81

Test 7, Performance UnderDivided Attention-ToneError, Minute 4 .79

Factor VII (Associative Factor VIII (MotorSpeed) Speed)

Test 4, Associative Learning Test 3, Performance UnderPart 1 -.60 Stress-Response Time .85Test 1, Kinesthetic Memory- Test 2, Perceptual Speed-Response Time .55 Reseponse Time .65

Test 4, Associative LearningPart 2 -.48

Factor IX (Manual Factor X (Manual.Dexterity I) Dexterity II)

Answer Sheet Markings, Wrongs .86 Large Tapping -.78Answer Sheet Markings, Rights -.39 Trace Tapping II -.71Test 2, Perceptual Speed- Answer Sheet Marking,Correct Answers -.35 Rights -.42

Discrimination-Reaction -.33Table Reading -.33

Factor XI (Perceptual' Speed)

Test 3, Performance Under Test 2, Perceptual Speed-Divided Attention-Perception Perception Time .72Time .84 Test 6, Concept Identi-

fication-CorrectAnswers -.33

aTaken from Hunter (1975).

14

Thus, examining the factors in Table 2-3 and Table 2-4 togeL,er, shows that: (a) five factors are predominantly identified by appara-tus tests (i. e., Visual Tracking, Auditory Tracking, Motor Speed, As-sociative Speed, and Perceptual Speed), (b) five factors were mostlyidentified with paper-and-pencil tests; Verbal, Spatial Relations, Me-chanical, and Manual Dexterity (I and I), and (c) one factor--FiguralMemory--was identified by both apparatus tests and paper-and-penciltests nearly equally. Hunter (1975) claimed such a factor configurationwas encouraging since it suggests that apparatus tests may make aunique and significant contribution (over paper-and-pencil tests) to theprediction of success and failure in technical training and work. Indeed,in a later study, Hunter, Maurelli, and Thompson (1977) showed sever-al of the same perceptual/psychomotor measures as predictive of per-formance in technical training schools.

Rarick and Dobbins' Taxonomy

Another perceptual/ psychomotor ability taxonomy, based on fac-tor analysis of a group of tests, was reported by Rarick and Dobbins(1975). As did Hunter, they described four perceptual/psychomotorability factors based on the factor loadings of each test on each factor(Table 2-5).

Noting that the Rarick and Dobbins (1975) taxonomy containsmore gross levels of abilities than does Hunter's (1975), the results ofany factor analytic study may be inferred to be rather situation specific.The number and types of tests employed, along with the sample of sub-jects used and conditions of testing, limit the scope and definition of thefactors eventually extracted. To increase comprehensiveness and gen-erality, additional factor analytic studies are needed which examine dif-ferent tests, different subject samples, and different test conditions.

Applied Psychological Services' Taxonomy

At Applied Psychological Services (Pfeiffer, Siegel, *aylor, &Shuler, 1978), a task taxonomy was developed to categorLze tasks onthe basis of their perceptual/psychomotor ability requirements. Thetaxonomy was developed with reference to military tasks and was delib-erately limited to categories about which there is considerable back-ground. The included abilities and their respective definitions are:

1. Vision--the ability to visually detect objects andrelations among objects such as movement or re-lative distances.

15

Li_ __ _

'rab1e 2'-

Tht H: i'' 1 ( i[ )ofntifl C ( p()iIjwn I t '0 i 1, 'M k I )mu 11vi I

an h' aliiabies t2hil 1~-'I wu'k illw AI~

AiH> , KiJI'-difl to i 1 l 1, alld I lI)lI1; (

Ii I I I T I

I o Ij)

hiiv 'ot diit~oi hlu'l-w ihl I& ' p iiiip ' 3 ii )-

hI i, ri II,

d'

2. Hearing--the ability to detect significant soundsamid competing sounds.

3. Strength--the ability to move objects using thebody and limbs.

4. Impulsion--the ability to react quickly to lightand sound by making explosive movements suchas tapping, running, and jumping.

5. Motor Speed--the ability to maintain a high per-sonal tempo and perform accurately using arms,hands, and fingers.

6. Static Precision--the ability to maintain goodbody balance and arm steadiness while aiming.

7. Dynamic Precision--the ability to maintain bodybalance and make accurate aiming movementswhile the body is in motion.

Miscellaneous: General Test Batteries

The final set of perceptual/psychomotor ability concepts review-ed here concerns those attributes measured within published generaltest batteries. Five major test batteries were identified as possessingrelevant perceptual/ psychomotor ability measures.

General Aptitude Test Battery (GATB)

'rhe GATB was developed by the United States Enployment Serv-ice. It was constructed primarily for vocational and employment coun-seling purposes, and includes 12 separately timed tests. The percCptualmotor ability factors included

1. Form Perception--the ability to perceive per -tinent detail in objects or in pictorial or graph-ic material; to make visual comparisons anddiscriminations and to see slight differencesin shapes and shadings of figures and widthsand lengths of lines.

2. Clerical Perception--the ability to perceivepertinent detail in verbal or tabular material;to observe differences in copy, to proofreadwords and numbers, to avoid perceptual er-rors in arithmetic computation.

17

3. Motor Coordination--the ability to coordinateeyes and hands or fingers rapidly and accurate-ly in making precise movements with speed; tomake a movement response accurately and quick-ly.

4. Finger Dexterity--the ability to move the fingers;to manipulate small objects with the fingers, rap-idly or accurately.

5. Manual Dexterity--the ability to move the handseasily and skillfully; to work with the hands inplacing and turning motions.

Differential Aptitude Tests (DAT)

The DAT is a widely used, general aptitude test battery. Unlikethe GATB, the DAT was developed more to measure vocational aptituIethan to aid in vocational placement. Consequently, the D)AI" measuresmany more cognitive and "intellectual" abilities than does the GATB.Of eight tests, only one appears to measure a perceptual or psychomo-tor factor: Clerical Speed and Accuracy--the ability to respond quickl 'and accurately to simple visual perceptual tasks.

Flanagan Aptitude Classifications Tests (FACT)

'The FACT was developed 'n tile basis of actual job behaviors andcritical work incidents. There are 14 so-called "job element" tests aim-ed primarily at semiskilled, skilled, and clerical functions. Eight testscan be taken as measures of Perceptual/psychomotor abilities:

1. Inspection--the ability to spot flaws or i-mperfc-tions in a series of articles quickly and accurate-ly. The test is designed to measure the type ofability required in inspecting finished or se mi-t'in-ished, mnanufactul'ed items.

2. Coding--the Sled and acLuracy of k'Odilig tYvpicaloffice infor mation. A high score r'an be obt'iinedeither by learni ng the codes quickliy o' bY speed illIcr forming a si mple clerical task.

3. NMe inory- -the ability to learln and rcclel l er the

codes given il fit(' coding test; the ability ,to Incmin-orize printed mate'rials.

18

4. Precision--speed and accuracy in making verysmall, circular, finger movements with onehand; and with both hands working together. Thetest samples the ability to do precision workwith small objects.

5. Scales--speed and accuracy in reading scales,graphs, and charts. The test measures scale-reading of the type required in engineering and

similar technical occupations.

6. Coordination- -the ability to coordinate hand andarm movements; to control movements in asmooth and accurate manner when these move-ments must be continually guided and readjustedin accordance with observations of their results.

7. Patterns--the ability to reproduce simple pat-tern outlines in a precise and accurate way.Part of the test requires the ability to sketch apattern as it would look if it were turned over.

8. Components--the ability to identify importantcomponent parts. The samples used are linedrawings and blueprint sketches. This perform-ance represents the ability to identify component,in other types of complex situations.

Employee Aptitude Survey (EAS)

As the name implies, the EAS was developed specifically for in-dustrial application, in particular, employee selection. 'rhe EAS pos-sesses 10 short, factor analytically derived tests--three of which canbe identified as measuring perceptual/psychomotor attributes:

1. Visual Pursuit--the ability to quickly and accu-rately trace lines visually through an entanglednetwork.

2. Visual Speed and Accuracy--the ability to seesmall details quickly and accurately, as requir-ed in visual inspection and clerical work.

3. Manual Speed and Accuracy--the ability to makefine finger mnovements rapidly and accurately.

- (

Guilford-Zimmerman Aptitude Survey

r'he Guilford-Zimmerman Aptitude Survey is the final test bat-tery considered. One of its seven tests appears to measure a perceptu-al/psychomotor ability: Perceptual Speed--the ability to see visual de-tails quickly and accurately by matching identical sketches of everydayobjects.

The perceptual/psychomotor measures within each of the five

batteries reviewed provide well defined attributes. The batteries pos-sess a strong research data base and, except for the FACT, define theirmeasures through factor analysis. As with Hunter (1975) and Rarickand Dobbins (1975), however, iney are not comprehensive taxonomicsystems, but represent a few selected dimensions. The perceptual abil-ities are operationally defined by the test batteries and lack, to some de-gree, conceptual meaning outside the tests themselves. In developing amore generally useful taxonomy, the value of the test measures rests inthe perceptual/psychomotor ability concepts that can be derived andadapted independent of the test batteries.

Suggested Taxoinomy for Determining the Perceptual/PsychomotorAbility Requirements in Selected Air Force Specialties

Taken together, the literature review yielded 89 perceptual/psy-chomotor abilities that are not unique. Added to this, a list of 17 per-ceptual/psychomotor abilities which are representative of the abilitiesrequired for task performance in Air Force specialties was provided aspart of the contract. Accordingly, a total of 106 perceptual/psychomo-tor abilities was on hand (Table 2-6), from which a final taxonomy wasdeveloped for employment in subsequent phases of the present work. Thetaxonomy was chosen through the following seven-step procedure:

1. Identical and apparently redundant abilities werecombined on this list; 31 items were thus com-bined.

2. Vaguely defined and grossly categorized abilitieswere eliminated from the list; two were elminated.

3. Abilities un'c lated to the perceptual/psychomotordomain and nonrepresentative of the ability require-ments in Air Force career fields were eliminated;12 were eliminated.

20

Table 2-6

List of PIeceptual/lsNychoinotoi Attributes

I Berliner (Rabideau, 1964)

1. Perceptual Processesb

2. Motor Processesb

I Fleishman (1966)

3. Control Precision4. Multi Limb Coordination5. Response Orientation6. Reaction Time7. Speed of Arm Movement8. Rate Control (Timing)9. Manual Dexteritya

10. Finger Dexterity11. Arm-Hand Steadiness12. Wrist-Finger Speed (Tapping)13. Aiming (Eye-Hand Coordination)14. Extent Flexibility15. Dynamic Flexibility16. Static Strength17. Dynamic Strength18. Trunk Strength19. Explosive Strength20. Gross Body Coordination

21. Gross Body Equilibrium22. Stamina (Cardiovascular Endurance)

III Harrow (1972)

A. 23. Segmental Reflexesc24. Enter-segmental Reflexesc25. Supra-segmental Reflexes

c

26. Locomotor Movements¢

27. Non-Locomotor Movementac28. Manipulative Movements

c

29. Kinesthetic Discrimination30- Visual Discrimination31. Auditory Discrimination32. Tactile Discrimination33. Coordinated Abilitiesc34. Endurance

a

35. Strength36. Flexibility37. Agility38. Simple Adaptive Skillc39. Cumpocnd Adaptive Skillc40. Complex Adaptive Skill"4L. Expressive Movementc42. Interpretive Movementc

21

Table 2-5 (cont.)

B. 43. Kinesthesisa44. Body Aw'areie~sl'.5. Visual Acuitv46. Visual Tracking47. Visual Memory48. Figmre Ground Discrimiaztion49. PerceptritJ Constaocy50. Audi to'cy A .uitry

31. Auditory TrackitF52. Auditc, em r[ory

IV Hv,tte (1975)

53. VJsoal l ,€k ir

14. Auditory TracT.qg55 Figural, Mewroy56. Position !rmory57. Motor Spee

a

58- Associpte Speed59. PercepLual Speed60. Gross ManAl Dexterirt'61. Fine Mattual De-:.rityt

V Rarick and Dobbins (1975)

62. Sute,gth-Powar-Body-Si zea63, Gross Body Coordinationa64. Fine Motor .Nbilitie.

- a

63. Balance3

VI Pfeiffer, Siegel, Taylor, and ScbleL (1978)

66. Vision67. Hearing68. Strength"69. Impulsion70. Motor Speed71. Static Precision77. Dynamic Precision

VII lHiscellaneouis: ('eaeral Test Baiarles

A. GATB

73. Form Perception

74. Clerical Perception75. Motor Coordination,76. Finger 'Dexteritya77. Manual Dexterity8

B. DAT

78. Clerical Speed and Accuracya

22

Trable 2-6 (cont.)

c. FACT

79. lnspectiona

80. Coding81. Precisiona8?. Scales83. Coordinationa84. Patterns85. Componentsa

D. EAS

86. Visual Pursuit87. Visual Speed and Accuracy

88. Manual Speed and Accuracya

E. Guilford Zimmerman Aptitude Survey

89. Perceptual Speeda

VIII USAF Representative Perceptual Motor Abilities

90. Multi Limb Coordinationa

91. Reaction Timea92. Control Precisiona

93. Rate Controla

94. Manual Dexteritya

95. Finger Dexterity5

96. Arm-Hand Steadinessa

97. Wrist-Finger Speeda98. Aiming99. Depth Flrception

100. Near Visual Acuity

101. Far Visual Acuity102. Size Perceptiou103. Sensory Acuity

104. Color Vision105. Lang Term Memory

106. Short Cerm Memory

Notes; * Hunter (1975) labels theee a!Manual Dexterity L and Xanu;;]DexLerity LI.

a Eliminated in Step 1 oi laxonomy

Derivation.b Eliminated in Step 2 of I.Ixonomy

Derivatio,c Eliainated in Stilp 3 ol 'laxonomyDerivation.

23

4. The remaining 61 abilities were rated on the fol-lowing eight criteria:

a. Compatibility--the ability should be fullycompatible with the Air Force task struc-ture.

b. Understandability- -the ability must bereadily apparent and comprehensible toAir Force users.

c. Objectivity- -the ability should allow forstandards of evaluation which are freefrom bias.

d. Scalability--the ability should allow forthe assignment of a magnitude value (anumber) to the tasks of a job relative tothe amount of the ability required for per-formance.

e. Validity--the ability should be based onacceptable constructs, relevant to AirForce job content, and seem reasonableto Air Force users.

f. Reliability--the ability should be amen-able to psychometrically reliable data ac-quisition methods.

g. Comprehensiveness--the ability shouldbe applicable to the full range of tasks in-volved in Air Force c,-reer fields.

h. Unidimensionality- -the ability should beunique.

Two other possible criteria--practicality and cost-effectiveness--were excluded since each directlyconcerns whole taxonomic systems rather than theindividual abilities.

Independent ratings of each remaining perceptual/psychomotor ability on each criterion were madeby two psychologically trained and experienced rat-ers who possessed knowledge of the different typesof Air Force career fields and the tasks performedin them. The ratings were made on a 5-point scale,where:

24

5 = the ability fully meets the criterion4 = the ability largely meets the criterion3 = the ability meets the criterion to a

moderate extent2 the ability meets the criterion mini-

mally1 the ability hardly meets the criterion

at all

Both the order of the abilities and the criteria onthe rating sheets were altered between the tworaters.

Although each ability was rated on eight criteria(488 ratings), a degree of variation which was use-ful for discrimination between abilities existed ononly two criteria: compatibility and comprehen-siveness. Accordingly, subsequent steps in selec-tion of abilities concerned only these criteria.

5. The independent ratings provided by the two raterson compatibility and comprehensiveness were com-

pared to determine the agreement and disagreementbetween the raters.

Agreement was defined as a difference between rat-ers of one scale point ot less, while disagreementwas defined as a difference of two or more scalepoints. All rating disagreements were resolved inconference by the two raters. There was a total of28 disagreements out of a possible 122 ratings (2 x61) and hence, 77% of the ratings were in agreementas defined.

6. The ratings were summed across the criteria ofcompatibility and comprehensiveness and two rat-ers for each ability (Table 2-7) to determine thecutoff points which included the top rated percep-tual abilities and top rated psychomotor abilities.

1 A perceptual/psychomotor ability was classified either as perceptualor psychomotor on the basis of its emphasis on stimulus or responseproperties. Those abilities more concerned with the stimuli and sen-sory systems were labeled perceptual; those more involved with theresponse were called psychomotor.

25

'[he cutoff point for' the perceptual abilities was 12,which included 16 abilities; for motor abilities thecutoff was 10, which included 17 abilities. 'Fable2-7 identifies the top-rated perceptual and psycho-motor abilities.

7. The final set of perceptual/psychomotor abilitieswas selected.

Both r'aters examined each of 33 top rated per'cep-tual/psychomotor abilities to deter mine their over'-all acceptability and desir'ability for inclusi,,n in thetaxonomy.

In the last step, the raters in conference examined the top-r'at-ed abilities, as well as the lower-rated abilities, and judged each. Fhisprocess led to the elimination of several top-rated abilities and tile ad-dition of a few lower-rated abilities. Several ground rules were impos-ed for application during this process:

1. Exclude the more physical, strength, and balanceoriented abilities.

2. Combine similarly defined and closely related abil-ities.

3. Include more specifically defined abilities in theplace of grosset and more broadly defined abil-ities.

Together, the judgments and decisions br'ought the total list ofabilities down to 13. These final 13 abilities and their respective defi-nitions are as follows:

1. Control Precision--the ability to perform rapid,precise, fine controlled adjustments by either arrnand hand movements ot' leg moverments.

2. Manual Dexter'ity--the ability to per'form skillful,well-dir'ected arm and hand movements to manipu-late either fairly lar.ge or fair'ly small objects un-dei, speeded conditions.

3. Finger' )extetrity--the abil 'itv to per'form skillful

manipulations or small objects with the finger's.

2 b

Table 2-7

Total Sum and Sums of the Compatibilityand Comprehensiveness Ratings by Rater

Rater 1 Rater 2 Total

Control Precisionb 7 7 14Multi Limb Coordinationb A 6 10Response Orientacionb 4 5 9Reaction Timeb 6 8 14Speed ot Arm Movementb 5 7 12Rate Control (Timing)b 4 6 t0Gross Manual Dexterityb 10 8 18Finger Dexterityb I 7 14Arm-Hand Steadiness 6 7 13Wrist-Finger Speed (Tapping)b 5 7 12Extent Flexibility 3 4 7Dynamic Flexibility 4 5 9Static Strengthb 5 5 10Trunk Strength 4 5 9Explosive Strengthb 5 5 t0Gross Body Coordinationb 7 7 14Cross Body Equilibriumb 5 5 10Dynamic Strengthb 5 6 ILStrength b 5 5 10FlexibilLty 5 6 U1Agility 5 4 9Kinesthetic Memory 4 6 10Visual Acuitya 8 8 16Visual-Tracking - 6 10Visual Memory 7 7 14Figure Ground Discrimination 6 4 10Perceptual Constancy 5 9Auditory Acuity 5 5 10Auditory Tracking 3 4 7Auditory Memory 4 4 8Figural Memory 6 5 11Position Menory, 4 5 9Motor Specdb 6 8 14Associate Speed 5 6 11.Ferceptual Specda b i 13Impulsion . 10Static Precision 5 5 10Dyamic Precis iot 3 1 6Vision' 8 7 15Hlearing" 8 6 14Form Pei Lep( ion

0 7 6. 13

Clerical Percepton3 6 6 12

Scales3

6 6 12

27

Table 2-7 (cont.)

Visual Pursuit 5 3 8Visual Speed & Accuracya 7 7 14

Depth Perception2

7 5 12

Near Visual Acuitya

9 8 17

Far Visual Acuitya

8 7 15

Size Perceptiona

7 8 15

Sensory Acuity 6 6 12Color Vision

a 7 7 14

Short Term Memorvb 8 7 15

Long Term Memoryb 7 7 14

Fine Manual Dexterity b 7 7 14

Patterns 5 5 10

Coding 4 5 9

Aiming 4 6 10

Kinesthetic Discrimination 4 6 10

Visual Discriminationa

8 8 16

Tactile Discrimination 5 6 11

Auditory Discriminationa

6 7 13

a top rated perceptual abilities.

b top rated psychomotor abilities.

28

4. Multilimb Coordination- -the ability to coordinatethe movements of a number of limbs simultaneous-ly, e. g., two hands, two feet, and hands and feettogether.

5. Rate Control (Tracking)--the ability to performcontinuous anticipatory motor adjustments relativeto changes in speed and direction of a continuouslymoving object.

6. Visual Speed and Accuracy--the ability to perceivesmall details quickly and accurately.

7. Visual Memory--the ability to recall and state ver-bally or recall and reproduce through writing anddrawings based on past visual experiences.

8. Position Memory--the ability to recall rapidly andaccurately the position of objects from past experi-ence.

9. Auditory Discrimination--the ability to discriminateand interpret sounds.

10. Auditory Memory--the ability to recognize and re-produce either verbally or in writing prior auditoryexperiences.

11. Clerical Perception--the ability to read or copyrapidly and accurately pertinent details in scales,graphs, or charts.

12. Perception of Size and Form--the ability to seeslight differences in the size and shape of objects.

13. Depth Perception--the ability to determine the posi-tion of objects in space and to perceive in three di-mensions.

2 9

Measurement Considerations

Because the present work sought to apply the derived taxonomyto specific Air Force specialties and tasks within specialties, it seem-ed necessary (a) to set into focus any measurement problems associat-ed with such application and (b) to derive a scaling method for quantify-ing the extent to which each skill (taxonomic category) is associatedwith each task.

'Theoretic Issues

"When you cannot measure it, when you cannot express it innumbers, your knowledge is of a meager and unsatisfactory kind." Thepositive application of Lord Kelvin's dictum has enabled the psychologi-cal sciences to progress from the stage of observation and classifica-tion to one of ever-increasing quantitativeness. But in the search for

quantitative rigor, the investigator has often found that, for "numbers"to be meaningful, there is a requirement for constructing a scale.

According to Ekman (1968), there are three main research prob-lems in modern psychophysics and scaling. They include stimulus-re-sponse relations, psychophysiological relations, and responsk.-responserelations.

Stimulus-response relations belong in the classical Fechneriantradition. The problem here is to establish the functional relation be-tween the stimulus (S) and the response (1). An example might be therelationship between judgment of elapsed time (11) and actual elapsedtime (S).

Psychophysiological relations have been studied less frequentlythan S-1 relations. An example of a psychophysiological relation is therelationship between an affective response and the corresponding activ-ity of the sensory nervous system.

Response-,'esponse relations concern the function al relationshipbetween subjective variables. In this case, only behavioral or isycho-logical responses to variables are measured. In facl, the physical cn n-tinuum against which tie psychological continuum in is compared may heunknown or not even exist as a meastiable Scale. In an early article,Guilford (19:39) contrasted the more co'tmmon S-Hj researc it with -l-research and pointed to the importance of establishing r'es; 'onse- it is

relationships. Relationships between Ire ived lask alt iilIJts ;]tt pert-ceived trainting requirements are examples of -H r,,seam'it. Wh,'on d -ing It-I research, tile development ()f such a sc ale ittw l,es: (a) (efititi)n

3 n

of a psychological attribute, (b) quantification of the attribute, and(c)es-tablishment of the relationship of the attribute with a second psychologi-cal dimension. Moreover, it is necessary not only to quantify the psy-chological attribute but also to specify the conditions under which thequantification may hold. Examples of perceived dimensions might beestimated training time required, number of components in a unit orsystem, ratio of satisfactory to unsatisfactory job performances, esti-mated time to perform a given repair or replacement, or, in the pre-sent case, the amount of a specific skill required to perform a task.

Scaling Classes

Regardless of the type of probleii involved, all scaling methodscan be placed into one of three classes. The classes are representedby scaling methods designed to produce an ordinal scale, interval scale,or ratio scale of measurement. With appropriate transformations andsome assumptions, an interval scale is produced by such methods aspaired comparison and lank order, and a ratio scale is produced di-rectly by the methods of magnitude estimation and constant sum.

Hepresentative Scaling Approaches

Por the purpose of assigning the extent to which various taxonom-ic categories are involved in Air Force tasks, at least five scaling ,netli-ods are possible.

1. The rank order miethod refers to a procedure inwhich stimuli are hierarchically arranged by anobserver alotl sone continuum. '[his method isdistinctive be cause all of the stimuli to be catego-rized are present for simultaneous observation.Application of the melhod results in a categoryscale or an interval scale (with transformation).

2. The paired comparison method refers to a pro-cedure in which the stimuli to be evaluated arepresented to an observer in all possible pairs.'[his method results in a category scale or, withthe statistical treatment, in an interval scale. Animportant difference between such category meth-ods as paired comparison and rank order is thatthe forier per mits the same category to be usedmore than once, whereas the latter may not.

: i

3. In category scaling, the judge assigns each stim-ulus to one of a number of discrete categorieswhich form a continuum; e.g., "always," "some-times," "never." Any number of categories maybe employed, but five or seven categories areused most frequently. The result is an ordinalscale which some treat as an interval scale.

4. Magnitude estimation refers to a procedure inwhich an observer makes a series of direct nu-merical estimates of subjective impressions.Each rater is presented with a series of stimuliand is asked to assign numbers proportional tothe apparent magnitude of the stimuli. One vari-ation of this method permits all stimuli to be pre-sent for simultaneous observation; another pre-sents the stimuli one at a time.

5. The constant sum method refers to a procedurein which all stimuli to be evaluated on a psycho-logical scale are presented to an observer in allpossible pairs. The task is to divide a total of100 points among the members of the set. Theimmediate numerical result of this procedure isa scale value for each stimulus and also the ra-tio of all possible pairs of stimuli to one another.

[he first three methods are indirect or category approaches toscaling, whereas the fourth and fifth methods are direct or magnitude

approaches. For a more complete treatment, the reader is referredto Guilford (1954) or Torgerson (1958).

Magnitude methods and category judgment methods seem to bethe mainstay of modern scaling applications. New methodological devel-opments have been few (Cliff, 197:3).

Types of Continua

Since the 1957 article of Stevens and Galanter, cont, mporary psY-chophysics has faced the embarrassing fact that the two classes of scal-ing procedures are often unable to produce the same scale of sensorymagnitude (Galanter & Messick, 1961). According to Stevens and G(Tlan-ter (1957), the relation between category and m agnitude scales is nonlin-ear for one class of perceptual continua, wheteas for" another class ofperceptual continua the relation may be lineam (p. :177).

:1) 2

Class I (prothetic) continua are characterized by a nonuniformityof discriminal sensitivity along the entire length of the continuum, i. e.,the subject is more sensitive to differences at tile low end of the scalethan at the high end. One would thus expect the just noticeable differencebetween stimuli to be greater at the upper end of tile scale than at the low-er end of the scale. Class I probably includes the dimensions that are or-dinarily called quantitative, e.g., length and weight (Stevens, 1961).

Class ii (metathetic) continua are characterized by a uniform-ity of discriminal sensitivity along the entire range of the continuum,i.e., there is uniformity of discrimination by an observer over the en-tire range. One would thus expect the just noticeable difference to re-0120 i approximately constant. Class II probably includes the dimensionstk< re ordinarily called qualitative, e.g., pitch and visual inclination(Siuven-., 1961) or job complexity (Pfeiffer & Siegel, 1966a).

For Class II (metathetic) continua, the category scale may be alinear funci i.on of the magnitude scale, whereas on Class I (prothetic)continua, or e scale should be a logarithmic function of the other scale(Galanter & Messick, 1961).

Relating One Scale to Another

There is also a pr(,Ilein when data derived from one scaling meth-od are related to those derive(i from another scaling method. Across-method comparisons do not at wvys yield a consistent set of results.

Generally, most studie., i'tggest that scales obtained by the cate-gory methods are usually logaritihnmic transformations of ratio estimationscales. According to early studies cited by Ekman and Sjoeberg (1965),these results have held true even for stimuli whose physical correlatesare quite complex. For example, Wliitlock (1963), who evaluated Jobperformance, found his data to satisfy one criterion for prothetic contin-ua. When he compared category and niuinitude methods on a continuumof apparent desirability of performanc, of e mployees, a concave down-ward curvilinear function was obtained.

More recent evidence has indicated a variety of relations betweenratio and category scales. While JIohn (1969) found the usual logarithmicrelation between category ratings and magnitude estimates of loudness,Sjoeberg (1968a, 1968b) found linear relations ainnng magnitude scalesand category scales of facial expressions when one lype of analysis wasperformed and curvilinear relations in another study. Gregson, Mitchell,Simmonds, and Wells (1969) found that, when two anchors were given forratio judgments of odors, the scale acted like a category .i,':

33

Montgomery (1975) summarized studies in which the importanceof methodological differences between category rating and magnitude es-

timation was examined. In one study, the form of a scale obtained by adirect estimation procedure was found to vary with: (a) the range of re-

sponses actually used by the subject, and (b) the amount of freedom inchoosing a highest number. When the category rating and a magnitudeestimation task were procedurally similar in both respects, the result-ing scales were linearly related to each other.

Gibson and Tomko (1972) showed that category and magnitudescales were linearly related when the end points of the category scalewere selected in such a way that they coincided with the previously de-termined range of magnitude estimates.

According to Montgomery (1977), given a constant Weber functionfor the category scale, the category scale is quite close to the Fechner

integral of the Weber function of the magnitude scale. From this, itmight be inferred that the general psychophysical differential equation isvalid for the relationship between category and magnitude scales. Fur-thermore, according to Montgomery, this relationship suggests that thecategory scale is a discrimination scale. Unfortunately, the assumptionof a constant Weber function for category scales is often contradicted byempirical results. The standard deviations for category ratings are typ-ically greatest in the middle and decrease toward both ends. Eisler andMontgomery (1974) suggested that this discrepancy between theory andempirical data could be explained in terms of bias or distortion in the

Weber function of the category scale. Apparently, extreme stimuli arediscriminated more easily than stimuli in the middle range. Clearly,additional data may be required to sort out the principles which deter-mine the kind of relation to expect between category and magnitudescales.

Siegel and Pfeiffer (1966b) attempted to sort out this relationship.They factor analyzed the results of a set of category and magnitude scal-

ings of technical job attributes. Full factorial congruency was not indi-cated. Siegel and Pfeiffer supported the use of the paired comparisonmethod over the rank order and the magnitude estimation methods. It

seems as if the judges in this study may have changed their frame of ref-erence and differentially emphasized discrimination, boundary mainte-nance, and perceptual organizational aspects when employing these lattermethods.

Eisler (1962, 1963) and Eisler and Montgomery (1974) summar-

ized some opposing viewpoints concerning the relationship between Inag-

nitude and category scale,-.

34

One view, as proposed by Stevens (1957), is that the categoryscale is, in effect, a distorted magnitude scale. According to this view,subjects who make magnitude judgments are influenced by the variationin ease of discrimination at the low and the high ends of the scale. Ifease of discrimination varies as a function of the position along the scale(low to high), then the function relating category to magnitude scaleswould likely be logarithmic. Thus, proponents of this view try to ex-plain away any differences so that the threat to a single psychophysicallaw is eliminated. Stevens (1962) suggested that apparent departures fromthe basic law might lead to a new and deeper understanding of sensorymetrics.

The opposing view regards the task of category rating as obvious-ly different from the tasks of ratio or magnitude estimation. Emphasisis placed on the methodological differences inherent in the category andmagnitude procedures. Proponents of this view suggest that the categorysituation yields a measure of the subjects' uncertainty whereas the mag-nitude method results in an estimate of subjective magnitude. Some ofthe variables given by other investigators to account for these differenceshave included size of subjective range (Engen & McBurney, 1964; Gibson& Tomko, 1972), spacing of stimuli (Pradham & Hoffman, 1963), the sub-jects' intent, ability to discriminate, and expectations (Stevens & Galant-er, 1957), distortion of the Weber function (Eisler & Montgomery, 1974),and response bias (Schneider, Parker, Valenti, Farrell, &Kanow, 1978).According to Montgomery (1975), these scaling methods differ in four re-spects:

1. Rule of assignment of responses to subjective magni-tudes. This factor denotes the instructions to judgesubjective ratios in magnitude estimation versus theinstructions to judge subjective differences (intervals)in category rating.

2. Openness of the response set. By openness is meantthe degree of freedom given to the subject to choose alowest and a highest number. In category rating thereis usually no openness, whereas in magnitude estina-tion the choice of a lowest and a highest number is leftto the subject.

3. Range of numbers. '[his factor denotes the range ofnumbers between the lowest and the highest tumberused by the subjects. Usually, the range of numbersused by the subject in magnitude estimation is muchwider than the range of numbers used in typical cat-egory rating,

35

4. Discrete versus continuous set of numbers. In cat-egory rating, the subject is usually allowed to useonly integers, whereas in magnitude estimation allpositive, rational numbers are permitted.

The results of Montgomery (1975) showed that the form of a scalevaries with: (a) the range of responses actually used by the subject, and(b) the subject's freedom of choosing a highest number as a response.The other factors that were investigated played only a minor role. Anarrow range and fixed upper scale value yielded the typical categoryscale, whereas a wide range with no restriction on the highest value pro-duced the typical magnitude scale.

Invariance When Scaling Job Requirements

A case in which the relationship between category and magnitudescales was found to be linear was demonstrated by Pfeiffer and Siegel(1966a). In their study, magnitude and category scaling methods were em-ployed by journeyman electronics personnel to scale the apparent complex-ity of various aspects of their own job. The resultant data indicated thatessentially equivalent scales were produced across the methods and thatthe continua of perceived complexity of four job activity stimuli and of 16electronic circuit stimuli were metathetic. This latter conclusion wasbased on the relative homogeneity of interindividual discriminal disper-sions for all the stimuli and also on the linear relation between the scalesresulting from category and magnitude scaling procedures. The absenceof large distortions as the result of the introduction of different methodssuggested support for a single psychophysical law in the avionics job per-formance area. Studies supporting scale invariance of this general typehave also been demonstrated in academic job areas. Pfeiffer (1970),drawing on techniques first developed at Applied Psychological Services,used both a magnitude estimation and a counting procedure to determinethe requirements of college professors. American and Suropean studentsand European professors were used as subjects. Correlations betweenthese scales, ranging between . 95 and . 99, indicated strong linear rela-tionships for all subject groups across scaling methods. The goodness offit obtained by linear procedures in conjuntion with supportive data on uni-form discriminal dispersions and the linear R-R relationships suggest ametathetic continuum foe" the range of values investigated. Moreover, theconsistency of these investigations about the continuum of job complexityin the technical job areas presented suggest that this type of R-R scalinganalysis of job complexity which included development of associated phys-ical correlates could be done well in advance of the normal system devel-opment cycle and by subjects with diverse backgrounds (Siegel & P'feiffer,19 6 6 a). In the Siegel and Pfeiffer (1966a) study, the relationship between

36

electronics job activity complexity, as perceived by maintenance techni-cians, and the scale value of these activities, as perceived by psycholo-gists and physicists, was investigated. Guilford's Structure-of-IntellectModel(Guilford, 1967) was employed as referent. Magnitude estimationmethods, based on the separate and independent judgments of technicians,physicists, and psychologists, were employed to derive the required da-ta. Moderate to fairly strong relationships were found between the scalevalues of the intellective factors involved in the job activities and the per-ceived complexity of the activities for all subject groups.

Implications

The present review possesses a number of implications foi- thestudy which aims to investigate perceptual/psychomotor factors in AirForce specialties. The fact that psychological scaling may not alwayswork perfectly should not be taken as prima facie evidence against theuse or invariance of such scales in general. Many cases of scale invar-iance have been reported- -particularly for metathetic continua (e. g., Ek-man & Kdnnapas, 1963; Pfeiffer & Siegel, 1966b). However, the infor-mation on hand suggests that such scaling should be performed cautious-ly ana that the sensitivity of the attribute being scaled to the type of scal-ing method should be determined prior to any major scaling endeavor.Such an investigation represents an early step in the present effort. If itis found that the scaling techniques are all, with minimal error, a linearfunction of one another, the selection of a scaling technique for employ-ment can be based on considerations that are other than methodological.If the customary cost and time criteria are employed, a categorical ormagnitude estimation method would appear superior. These methodsyield values which can be used directly.

A persistently aggravating problem in scaling is posed by the factthat each judgment is affected by the other stimuli being judged. Pairwiseratio judgments and rankings seem most likely to show such effects. Un-fortunately, context effects have also been shown to occur when categoryscales are employed (Cliff, 1973).

A strategy which has considerable historical support in scientificpractice is to say that the preferred method is the one that is most direct.Magnitude and categorical scaling have this virtue. The only assumptionis that the observer is able to carry out the instructions to quantify per-ceptions. Alternatively stated, this means that the subject reacts to num-ber stimulation in the same way as to any stimulation.

Finally, one comforting fact is thnt, regardless of iheoretic is-sues, psychological scaling of job tasks on attributes has been shown to

37

yield useful, reliable data. In a study completed for the American Tele-phone and Telegraph Company (Siegel & Federman, 1976), a variety ofcraft and clerical tasks were rated by job incumbents on a number of lax-onomic attributes. Adequate between-rater reliability was achieved andthe data were able to provide a basis for job evaluative techniques.

I

III. METHOD DEVELOPMENT AND PRETEST

With the ability taxonomy described in Section II on hand, a meth-od was sought that would allow application of the taxonomy to a range ofAir Force career fields so as to determine whether each of these abilitiesis involved in the job performance of these career fields and if involved,the extent and influence of the involvement. Accordingly, some type ofjob analysis based on the derived taxonomny seemed indicated.

Methods of Job Analysis

Job analyses have been conducted, over the years, using a varie-ty of data collection procedures. The relative desirability of the avail-able methods depends on a number of factors. Perhaps the most impor-tant consideration is the goal of the analysis. At the general level, thegoal of any job analysis is to derive information which describes what isdone on the job, to develop information relative to the personal attributesrequired fo- job performance, or to derive a fuller understanding of thejob itself. The derived information may be qualitative or quantitative; theinformation may be molecular or molar.

Direct Observation by Analysis

Job analytic data may be collected while the analyst is directly ob-serving performance by job incumbents. This technique may appear op-timal in that it affords the opportunity to question the job incumbent at anystage of the work. It allows photographs and physical and temporal meas-ures to be taken as desired, and it may permit the analyst to performsome or all of the job. Data collected in this way are likely to be objec-tive and accurate, since the method does not rely on the job incumbent'smemory or expressive ability. Indeed, direct observation is often con-sidered the superior method for collecting job analytic data (Yoder et al.1958; Bechtoldt, 1951). However, the method is very costly in terms ofanalyst and job incumbent time, if all portions of a job are to be analyzed.The presence of the analyst may strongly influence the behavior of theworker, reducing the validity of collected data, and the results of apply-ing the method can be influenced by the analyst's perceptivity. The tech-nique may be applied only to analysis of a job which is currently beingperformed; actual equipment and sufficiently experienced personnel arerequired, as is the continued presence of a perceptive job analyst. Final-ly, the technique is suited to studies of the overt, obse rvable aspects ofjob performance, but may not be effective for analyses of less observable,underlying variables associated with job performance such as are pre-sently under consider ation.

39

Interview Methods

.Job analytic data are often collected through interviews with:single Job inc umbents, groups of job incumbents, or with others as-sumed to be expert with respect to tie job of interest. The group inter-view approach is also sometimes called the technical conference tech-nique. In interview techniques, data are generally collected away fromthe Job site. The analyst will usually follow a structured or semistruc-tured interview protocol. '[he goal of the interview may be to collect da-ta descriptive ()f the job itself, or to collect data concerning some under-lying variable or set of variables, such as required aptitude, skill, train-ing requirements, or work pace. Interview techniques may be expectedto be relatively inexpensive to apply, and they may yield a variety of jobrelated data. Group interviews allow immediate review of data accuracyby the entire group, and the data may be collected by interview when theapplicable equipment or facilities are unavailable or not in existence. Onthe negative side, accuracy and completeness of collected data are affect-ed by the skill of the interviewer and the capabilities of interviewees tocommunicate. Data collected concerning time spent on tasks, levels ofdifficulty, skills required, etc., may be affected by subjective bias orimperfect memory on the part of the interviewee. Data obtained from su-pervisors or experts may tend to reflect expectations of those personsand the image they wish to present, rather than actual experience or be-havior of job incumbents. Additionally, such data are likely to be color-ed by the range of skills and experiences held by the expert interviewee(s).'[he interview allows the acquisition of respondent insights beyond a sim-ple categorical reply. It allows the respondents to elaborate more fullyon their responses and to supply their own points of view. It also allowsthe interviewer to probe and to follow up on areas of doubt, concern, orambiguity. Such an opportunity can only be minimally provided in othermethods. However, the interview is an interpersonal interactive situa-tion and, as such, is subject ;o the influence of such situations. More-over, interviews are time consuming, and open-ended responses are sub-ject to interpretive vagaries. The qualitative data emerging from in-terviews are often more difficult to treat and, if content analytic meth-ods are involved, an additional er.ror source may be introduced.

40

Questionnaires and Task Lists

Collection of job analytic data through questionnaires is appeal-ing due to its economy in reaching a wide number of incumbents/super -

visors. Very little time is required of the job analyst for collection of thedata. However, the analyst has no control over and little idea of the at-tention or understanding applied in completing the questionnaire. Assuch, job analytic questionnaires are subject to the same biases as anyquestionnaire. The results may be biased by the wording of tihe set ofquestions employed or by their form. The response rate may be low andproblems may arise in analyzing obtained data.

The task list approach is a variant of the questionnaire approach.In the task list approach to job analysis, a list of the tasks performed onthe job is presented to the job incumbents or their supervisors. They areasked to complete a structured set of ratings relative to each task. Theratings may involve how frequently each task is performed, how hard itis to learn to perform each task, how serious the consequences of inade-quate performance of each task are, how complex each task is, the lengthof time between training and performance of each task, how long it takesto perform each task, how important each task is, or whether or not eachtask is performed in an emergency.

The technique has received extensive investigation and applica-tion within the Air Force. The technique has also been adapted by atleast one major business corporation, where it has been emploved fortraining requirements and organizational structure development purposes(Siegel & Federman, 1976). 'The method has also been adapted for im-plementation on a commercial basis by at least one industrial consultingorganization (Lopez, 1978).

Use of the method rests on the availability of a task list. '['he de-velopment of such a list depends on one or several of the job analyticmethods just described. However, if such a list is available, the task listapproach is comprehensive and economical, and the results are amena-ble to a variety of standard statistical manipulations. If the list is admnin-istered through the mail, there is little control over the quality of the rIe-sponses, the diligence of the respondents, ori the conditions under whichthe form is completed. However', if the lists are taken to tihe job in timun-bents, these problems may be avoided or controlled thirouh orientation-al training, administrator diligence, and standardization of admintistr a-tive conditions.

41

Daily Diary

In tile daily diary approach to job analysis, workers are asked tokeel) detailed records of their daily activities. This method is inexpen-sive and may provide accurate data. The obtained data are usually in aform which is not readily amenable to statistical analysis. This methodseems best suited only to determination of the tasks composing a job andthe time devoted to each. Unless the workers complete the diaries dur-

ing tihe course of their work, they may forget some details. And, com-

)leti sCh a diary during the course of the work may interfere with thework itself. Man blue collar workers do not feel at ease with this mneth-

od because they are unaccustomed to any written expositional task.

Critical Incident

In tihe critical incident approach to job analysis, observers record01' lItubntL II s ate asked to descr'ibe tasks or incidents which fall at theextes l oef01 some co'ltinuum such as task perforimalce time, d gl'ee ofphysical dllnand , or degree of risk. This technique is not useful for de -\elOlpit.g a dtailed description of a job, factors underlying a job, or re-q ui ted ai)I 01S ' rt job per'ormance. Hlowever, tie tech nique may pro-

vidt l u a ih Cr rt) 'te 0toltetida tio is for job mnodification or redesigil, be-

cause the data eli,.it ed by the technique will direct attention to those

aspects o1 a jot) whicl are extreme. The benefits and disadvantages gen-eorally a)plic.able to interview and observation or questionnaire tec hniques

of Job attalysis , as described above, apply when they are used within the

critical itcidetit ahiliroach.

4*2

Comparison of Various Approaches

The merit of each described job analytic procedure relative tovarious characteristics is summarized below:

-0

0m

07,4 7

.7,

0 0

- -. >. .,- .,

-0 Total

Direct Observation 4 + 0 4 1

Interview Mthods -4 1 0

Task List/ Questiotinaire -f 4 + 0 .5- -

Daily Diury 1 4 4 0 1 0 A 44

Critical Incidenit t' - - . .. 3-

The agl)gaic sums nay t)4, used IA) suggest lOUghll*\ an 01(0r of On"JJl

for the vrious tvuhiqui(tes. Of (0U1stn a& tuhniquv Iflt)Q W)1 svAnoi "Wth

(O siJ(arlti()fl gIV(l giv thet.p " ()lo lto nol'alii1 desired, ille g)l(s) of the

Spe, itic ala ,sis, thle ino muli alread", available, and the o sit inlts

Of A(e situatloll.

41

Task List Method

In the above analysis, thle task list method appeared to be a pre-ferredt method of approach within thle present stud. The mlethod receiv-ed thle higyhest absolute score. The task list method is hlighl\ applicableas a technique for deriving underlying abilities--a fundamental requ ire -nient of the present w ork. Mm eov e i, thle melhod waus c-onsideredi to becost effective, would interfere minimal ly wvith ongoing work, and wouldbe comprehensive, quantitative, and objective. R~elatilye to thle presenltprog~ram, thle task list method see med to possess thle foil owing adva n-tages:

1. timeliness--the method seemedt to allow achiey e-ment of thle r-equired goals within the requiredtime p)eriod.

2.compatibility -- the met hod is (0 rpatible wkit i current Air [Vorce practice.

3. balance -- the m eth od seemedl to allow iV.U lil (oil -

sideration of all taxonomic categories withI no cat -egorv being undulY emrpha-si/. d at the expense of'other categories.

DiscuIssion of Trask List Method

Thie task list method of job) anialv makes 'A t t as su mp tions

other thanii os e nor mallv, as su med for all n( ti s ill . As so il in al

understandable task list and set of' insil't ions, tilt' Inll1od also a-SIIllm

that thle per'son coniplJeting the form is knowleudgeable- about t~n' jot) andican accui-ately report that knowl'edge. Validit 'N cocfiiclollt roveati' lt orepJor'tll1 Wi aC c u aC V v found by iee andR ]usc 11 )t he ad equate -

IY high.

Anyonie using anl already dIeveloped List, inn st akt'p e11Ii le ciol Oc ~-hlenSiV eneQ 55Of thle list anld aS someC tilie job) has illl c awit'.a (d n 1 .1ig Ill leii-lerval between initial list developmnlt all , rillt applik at ionl. Suchi listsalso a ssu me that tile See I lial o iderli g of If!( eisa 15 IoI a s 1,i 1,l Uii 11an

factor affect ing tile resultant dlata und that treis l0It-, if'l. an sequctlt -

rotation of' lite task list sequee ;A( l'5 rateis;.

The task list approach has been h.hl\ l \ the All I (Wo ,Inl slin-vfeYs of a numbher ot' task chiaracterist It s, ii luili~ i f1'JtLiok ott tas t

lorimante," alliouit otf supeivis loll reljlili od m:ti ise-d, lask ( tmphlexit

I4

required for effective performance, importance of task to unit mission,etc. (Morsh, 1964). The Air Force has found that task lists may be ad-ministered economically. According to Morsh, test-retest reliability ofsuch lists administered by mail ranges around .70. Also, initial studiesof validity show little disagreement between data provided by job incum-bents and by their supervisors.

Within the Air Force context, Christal (1974) found the task listapproach to job analysis to be economical, and to be highly quantifiable.Additionally, the approach is held to yield data which may be readilystored, manipulated, analyzed, and reported by computer, and are amen-able to tests of validity and stability through standard statistical methods.

The method generally assumes that the rater will be free from bi-as. For example, a bias will result if a rater rates dishonestly. Theremay be raters who are openly dishonest or hostile to the rating procedureor job being evaluated. Alternatively, the rater may be uncommitted tothe rating task. Campbell, Durnette, Lawler, and Weick (1970) suggest-ed that lack of rater commitment is the most serious source of rating bi-as. They reasoned:

The most serious source of difficulty (bias) is a veryfundamental one - stemming from a common tendencyfor psychologists to impose their own beliefs about jobbehavior and their own systems for recording it uponthe persons whose task it is to observe that behavior...(It is) a lack of understanding and a lack of commit-ment to the observational (rating) task on the part ofobservers. As a consequence, they (the observers)tend to fill in the forms (job behavior, rating scales)with little conviction; the records contain large andfor, the most part inestimable errors (p. 118-119).

Error due to lists provided and error due to the configuration ofthe ability structure selected represent stimulus error. There is nodoubt that a stimulus task list, which is not objective, unambiguous, orcomplete or which otherwise does not incorporate the characteristics ofa well-designed, coherent set of items, will be less than fully useful.However, there is a considerable amount of information available abouthow to construct such items, how to present them, andi how to p!1.SfslIclear instructions to the user. These sources of er'or were expected lobe controlled in the present work by sufficient c-are and attention in laskitem writing and through clear rater instructions and training.

45

..... Gap.

Similarly, the error introduced because of the configuration ofthe selected ability structure was to be reduced by providing a meaning-ful ability taxonomy and further reduced through an adequate set of defi-nitions, examples, and instructions.

Pretest Sample and Task Sample

After the task list approach to data collection was tentatively se-lected, it had to be verified before employing it within a major data col-lection effort. Accordingly, a pretest was established. The specific goalsof this pretest were as follows:

1. Identify problems inherent in the anticipated pro-cedures, the instruments, and the measurenenttechniques.

2. Examine relationships between rater experience(i. e., supervisor versus subordinate ratings) andobtained information.

3. Assess the interrater reliaoility when employingthe method.

4. Measure consistency of responses (test-retest re-liability).

5. Serve as a test bed for determining time require-ments.

6. Determine the sensitivity of the obtained data tovarious methods of scaling.

Pretest Sample

To accomplish the pretest, two career fields, Fire P rotection(571XX) and Munitions Maintenance (461XX) were selected. The conisid-erations bchind the selection of these career fields were that a suffic ientnumber of job incumbents seemed available at each of two air bases andthat each career field represented a different aptitude area (as used b*"the Air Force for classification purposes). Fire Protection is classifiedin the general aptitude area, while Munitions Maintenance is a mechani-cal aptitude area specialty.

Thet job inUmb'nts ii the )wetest were all assigled to lEglill andHomestead Air l.'ot'ke lases. Sulpe'visors, i. e. , pay 'gad(es of" EI-(; orabove, anid suio'didaltis, I. v. , pay gi-ades of E'-5 and below, \eel' sam-pled. Tabh 3-1 (hj,4eSJ1. ,s 11.h satIph, by play g adc and c-areer field.

46

Table 3-1

Description of Pretest Sample

Pay Grade

Career Field E-5 or Below E-6 or Above Total

Fire Protection (571XX) 24 23 47

Munitions Maintenance (461XX) 22 24 46

Total 46 47 93

Task Sample

In an effort to develop pretest task lists that were fully repre-sentative of the tasks and duties performed by the personnel in the twocareer fields, Air Force Occupational Survey Reports (OSRs) were con-sulted, and tasks were selected in accordance with the following task/duty* sampling scheme:

1. Eliminate tasks and duties involving only super-vision, planning, and training.

2. Eliminate duties performed only by 25 percentor less of the members of a career field.

3. Establish a target list length of 20 tasks for eachcareer field. Weight the duties in each careerfield in accordance with the following weightingscheme:

% Performing Weight

75 or more 4

Between 50 and 74 3

Between 26 and 49 2

* A duty is a set of related tasks. Tasks are grouped under duties in an

OSR1.47

Sum the weights over the duties included andcalculate the ratio of the weighted sum in eachcategory to total task sample size, e. g., if thesum of weights over all duties is 20 and therea-e two duties with weights of four, then 40percent of the total number of tasks selectedwould be drawn from the duties with the weightof four (4 + 4 = 8; 8/20 = 40%). If the calcula-tion results in a fraction, add a task to the dutywith the largest number of tasks, within thegroup of duties with the same weight.

Trable 3-2 presents the results of the sampling procedure for the two pre-test career fields.

Development of Forms

A set of data collection forms was developed to allow the collec-tUon of information about the involvement of each of the 13 perceptual/psy-chomotor abilities in each task. The forms were prepared separately for,each career field but the forms for the two career fields were similar informat. Each form contained a full set of instructions and examples ofhow to complete the form.

Perceptual/ Psychomotor Abilities

The 13 abilities contained in the perceptual/psychomotor abilitytaxonomy were prepared for employment in the pretest. Several modifi-cations were introduced into the prior definitions in order to nake themmore meaningful to the anticipated job incumbent raters. Additionally,examples of activities involving each ability were developed for each de-finition. For each ability, two examples were developed to typify tasksin which a "high" amount of the ability is required, and two other exam-ples of activities were developed for which a "low" amount of the abilityis re~juired. The perceptual/psychomotor ability definitions list used inthe pretest, along with the examples, follows:

1. Finger Dexterity (FD)--skillful, coordinated, precise finger move-ments that involve the use of one or more fingers to achieve quickand accurate manipulation, insertion, or grasping of small objects.

Examples:

High i. Typing requires rapid movement of severalfingers to perform a sustained, coordinatedac2tivity.

48

Table 3-2

'[ask Sampling Procedures for Pretest Career Fields

Munitions Maintenance (461XX)

Duty Sample % Performing Weight * No. of Tasks **

C 95 4 3E 91 4 3F 49 2 2G 27 2 1H 67 3 3I 64 3 3J 33 2 1M 60 3 2N 48 2 2

Z= 25

* Ratio of weight to sample size; 4: 8/25 = 32%, 3: 9/25 36%,2: 8/25 - 32%

•* Number of tasks to be sampled from each weighted category;4: 32% x 20 = 6, 3: 36% x 20 - 8, 2: 32% x 20 = 6

Fire Protection (571XX)

Duty Sample % Performing Weight * No. of Tasks **

C+ 37 2 -

E 62 3 1F 83 4 2G 55 3 1H 53 3 1I 63 3 1J 71 3 2K 71 3 2L 70 3 2M 66 3 1N 79 4 10 82 4 2P 30 2 1Q 47 2 1R+ 28 2 -

S 35 2 2E =46

+ The OSRs did not contain task related data for Duties C and R due tovery low percentage data. Replacement tasks were selected from other

duties in the same weight category (Duty Q, by this procedure, should

have had two tasks sampled but the OSR reported only one task)

• Ratio of weights to sample size; 4: 12/46 = 26%, 3: 24/46 - 52%,

2: 10/46 - 22%

•* Number of tasks to be sampled from each weighted category;

4: 26% x 20 - 5, 3: 52% x 20 - 11, 2: 22% x 20 - 4

49

2. Painting, drawing, and lettering require theaccurate manipulation of an implement.

Low 1. Pulling the trigger on a revolver requireslittle finger dexterity because only one fin-ger is used in a relatively uncoordinated ac-tion.

2. Activating a light switch is low on fingerdexterity because only a finger and thumb(or finger alone) are used to throw the switchand no precision is required to position theswitch.

2. Manual Dexterity (MD)--skillful, well-directed, coordinated arm andhand movements to manipulate objects quickly and accurately (but,not controlling a machine).

Examples:

High 1. Assembling a radio because parts and toolsmust be manipulated accurately, carefully,and in a coordinated manner.

2. Welding a patch in an aircraft's skin requiresa high degree of skillful arm and hand manip-ulation.

Low 1. Closing a door is low in manual dexterity be-cause there is little directed activity or coor-dination involved.

2. Grasping pliers to hold an object requires somemanual dexterity, but little skilled movementsor coordination of the arm and hand.

3. Control Precision (CP)--rapid, precise adjustments by an arm, hand,(individually or simultaneously) to a machine's control mechanism(e.g., levers, pedals). The adjustments do not involve objects (e.g.,pencils, tools, electronic parts).

Examples:

High 1. Fine-tuning a radio dial requires control pre-cision.

2. Manipulating the gas pedal in an automobilerequires control precision because a car ishighly sensitive to slight changes in the pres-sure applied.

50

Low 1. Operating an on-off switch requires minimumprecision in going from one position to theother.

2. Activating the high beamn lights of a car withthe foot control requires a low level since itis accomplished with a single depression andno adjustments or precision are required.

4. Rate Control (Tracking) (RC)- -continuous and accurate arm, hand,or leg control adjustments to changes in the speed and/or direc-tion of continuously moving objects. The purpose is to intercept,control, or follow a moving object.

Examples:

H igh 1 . Tracking a target on a cathode ray tube (CRT)by keeping the target inside a cursor (circle)requires fine control adjustments to quicklymoving targets that move in several dimen-sions.

2. Auto driving requires continuous within-tol-erance adjustments of the steering wheel,

Low 1. Illuminating a slowly moving object with aflashlight requires little rate control be-cause the speed of movement is low and theperformance tolerances are high.

2. Walking with others requires a low amountbecause speed will vary slightly and the move-ment is in a readily predictable direction.

5. Visual Memory (VM)- -recall of things which have been seen in thepast and expressing the recalled visual information.

Examples:

High 1. Identifying a needed spare part in a group ofparts on the basis of appearance requires theability to recognize the spare part and specifyit by name and/or number.

2. Recalling the appearance of a given aircrafttype requires memory for visual information.

Low 1. Entering parts replaced on a maintenance formrequires a low level because the appearance ofthe parts does not have to be remembered or ex-pressed.

51

2. Scheduling work assignments requires a lowlevel because the appearance of the personsto be assigned does not have to be remem-bered.

6. Visual Speed and Accuracy (VSA)--seeing small, fine details quick-ly and accurately. It includes seeing differences in size and shape.

Examples:

High 1. Inspecting a part for rust, chips, mars,scratches, or marks requires the abilityto see fine details quickly. No recall isinvolved.

2. Examining and separating fingerprints quick-ly requires the ability to distinguish size andshape of fine details in a limited time period.

Low 1. Washing an aircraft because the visual ac-tions to be performed are quite gross.

2. Inspecting cargo for shipping size restric-tions requires a low amount because the de-tails are gross and few time restrictionsexist.

7. Position Memory (PM)--recalling rapidly and accurately the positionof objects from past experience. Emphasis is on recalling the posi-tion or location of where objects belong without having to express it.

Exa.nples:

High 1. Performing a preflight inspection becausethe location of the items to be inspectedmust be recalled.

2. Locating a part in an aircraft as the resultof a prior experience with the part.

Low 1. Fueling an aircraft because the fuelingpoints are quite obvious.

2. Filing requires a low level of position mem-ory because the alphabetical sequencing isobvious.

52

d

8. Auditory Discrimination (AD)--distinguishing sounds and interpretingthem.

Examples:

High 1. Detecting and determining a malfunction ina motor on the basis of the sounds of themotor.

2. Interpreting Morse code.

Low 1. Listening to a pipe to determine if water isflowing through it involves a low level ofauditory discrimination because the soundis distinctive and the discrimination is gross.

2. Determining whether or not a motor is run-ning requires a low level since there are on-ly two choices, on or off, and the sound of arunning motor is quite identifiable.

9. Auditory Memory (AM)--remembering, recognizing, and reproducingthe characteristics of sounds. This ability does not involve interpre-tation of the sounds.

Examples:

High 1. Recognizing a navigational tone in the pre-sence of conflicting signals and describingthe sounds verbally.

2. Explaining the characteristics and distin-guishing features of an emergency signal.

Low 1. Describing radio static on a maintenanceform.

2. Identifying thunder during an electricalstorm.

10. Clerical Perception (CLP)--rapid and accurate reading or copying of

details in scales, graphs, charts, or tables.

Examples:

High 1. Copying a long list of numbers from a tele-phone directory.

2. Obtaining information from a detailed partsreplacement stock list requires a high level.

53

Low 1. Standing guard duty at the base entrance andnoting numbers from automobile license plates.

2. Reading organizational charts.

11. Depth Perception (DP)- -determining positional relationships amongobjects in space.

Examples:

High 1. Operating an aircraft in a congested area re-quires a high amount due to the involved ar-rangement of the aircraft in space.

2. Flying in formation

Low 1. Loading an aircraft with cargo.

2. Attaching an auxiliary power unit (A1lU) toan aircraft.

12. Divided Attention (DA)--receiving and using information from morethan one source at the same time.

Examples:

High 1. Monitoring and directing aircraft in the con-trol tower based on information received fromthe radio and the radar requires a high levelbecause of the simultaneous use and integra-tion of complex information from more thanone channel.

2. Flying an aircraft under instrument flight reg-ulations (IFR) conditions because various in-struments must be read, a radio must be mon-itored, and the aircraft controlled at the sametime.

Low t. 'resting an electronic equipment on the basis ofa meter and a sound signal requires a low levelbecause although two channels are involved, theoperator can shift easily from one to the other.

2. Supervising several subordinates requires a lowlevel because the supervisor is not dependent on

, receiving the information simu~ltaneously.

54

13. Kinesthetic Memory (KM)--manipulating objects without benefit ofvisual guides or indications.

Examples:

High 1. Replacing screws in places that are not insight,such as under a dashboard, requires a highamount of the ability since the actions cannot beobserved and a blind positioning action is involv-ed.

2. Attaching a fitting to the pipe under a sink re-quires a high level because visual guides are notavailable during the manipulations.

Low 1. Screwing a light bulb into a socket in a dark roomrequires a low amount of kinesthetic memory be-cause the relationship is quite obvious.

2. Inserting a key into a keyhole in a dark room re-quires a low amount because the manipulationof the object is limited. Only an insertion is in-volved.

Perceptual/Psychomotor Ability Requirements Questionnaires

To provide a basis for- achieving the overall pretest goals--evalu-ation of the task list questionnaire method of data acquisition and the util-ity of the taxonomy--two data types were relied on: (a) how much eachability is involved in the performance of each task (amount) and (b) thevariability in the quality of task performance as a function of each specif-ic ability (performance quality variability). The first data type providesa measure of the relative saturation of a perceptual/psychomotor abilityin the performance of a task, and by summary in the performance of thecareer field. The second data type provides an indication of whether ornot the ability separates good from poor task performers.

Each data collection instrument was divided into two sections.The first was called "Amount of Perceptual/Psychomotor Ability"; theother was titled "Performance Quality Variability as a Function of Per-ceptual/ Psychomotor Ability."

Each section was further divided into two parts--one part requir-ed a categorical response while the second required a magnitude estima-tion. The survey respondefnts, thereby, yielded an evaluation of eachtask twice in each section of the form--once using a category scale andagain using a magnitude estimation scale.

55

For the categorical estimations, a five-point category scale wasused. The range of categories, in the amount section, was from "I"(very little) through "5" (very much), with the scale value of "3" (mod-erate) anchoring the midpoint. The scale appeared on a card which therespondents referred to while completing the form and also in the ap-propriate set of instructions. Two other qualitative responses were in-cluded: N = not performed in your squadron and NR = performed, butability not required in task performance.

The magnitude estimation scale provided greater freedom of re-sponse. rhe scale ranged from 0 to 100, and allowed the option of se-lecting any intermediate value. Qualitative descriptive anchors wereplaced under the scale of values at five levels. 'The instructions to thetwo sections and the two parts within each section of each questionnaireare presented in Appendix A to this report.

Pretest Administration

An Applied Psychological Services staff member, who was in-volved in their preparation, administered the forms at the two AirForce bases involved. The survey sessions were conducted in classroomsituations. About 12 to 25 respondents participated in each data collec-tion session.

r'he person who performed the form administration was experi-enced in data collection techniques in the military and was knowledgeableof the perceptual/psychomotor taxonomy and the job analytic concepts.

Administrator's Training

A set of administrator instructions was prepared for and review-ed by the administrator prior to his air base visits. Nhe instructionspresented a standardized procedure for the administration. All perti-nent procedural steps were outlined in the instructions. This allowed theadministration to proceed in a consistent manner across sessions.

The administrator instructions suggested the points to cover in abriefing period which preceded the respondents' completion of the forms.Additionally, the instructions suggested an administrative style and atti-tude which would be both helpful and motivating to the respondents.

56

Administrator Instructions

1. Assemble the group

2. Distribute the following

* forms by AI,'SC number (right corner of cover page)

* definitions list

* pencils

3. Introduction

* your name

" give company name and geographic location

" indicate that Applied Psychological Services is undercontract to the Air Force Human hesources Labora-tory, Brooks Air F'orce Base, to conduct this researchprogram

4. Purpose of research program

* to evaluate the abilities required in different Air 'orcejobs (i. e., the sensory and manipulative aspects of thejob, the things people do, and the performance aspectsof the job)

" the data will be used specifically to help the Air I'orcedetermine job performance requirements

* the global purpose is to improve the entire career de-velopment program in the Air Force

5. Why respondents were chosen for the study

* the best way to obtain the needed information is to comedirectly to the people who are most knowledgeable aboutthe jobs, namely, you the job holder

* you were chosen because, due to your job knowledge,your opinions of the amount of involvement of each -killin the performance of various tasks and the differencein the quality of performance (as a function of each abil-ity) will be more valid than the opinions of people whoare not as involved in the specialty

6. Explain the respondent's task

* there are 13 abilities that we are interested in

" the abilities, which are defined in the definitions listyou were given, appear across the tot) of the forn(demonstrate)

57

" there are a total of 60 tasks, 20 on each of these pages- -

the tasks appear vertically down the left side of eachpage (demonstrate).

* the tasks appearing in the form represent a small sam-ple of the tasks done on the job. They were obtainedfrom the Occupational Survey Reports (OSRs) of eachcareer field.

* there are two sections in the form. In the first section,you are asked to give your opinion of how much you thinkeach ability is involved in th performance of each task.In the second section, you are asked to judge how thequality of performance amoung job incumbents varies asa function of each ability. Each section has a set of in-structions and examples which will explain the purposeand method of responding.

" since this is a pretest of the survey forms and one pur-pose is to find out which one scale, of two different types,will be better to use, you will be asked to make the sameevaluations twice in each section. One of the scales,which is called a category scale (demonstrate) has valuesfrom 1 to 5. Enter the value that best describes youropinion. If a task is not performed in your squadron, thenyou would enter '"N" in the appropriate box (demonstrate).If you are of the opinion that a particular skill is not re-quired in the performance of a task, then you enter"N1- I"in the appropriate box (demonstrate). The other scaleyou will use is called a magnitude estimation scale (demn-onstrate). In using this scale, you may select any valuefrom zero to 100 that best describes your opinion. Makeevery effort to enter a judgment, even if you don't performthe task yourself and never did. We are not interested inwhat you personally do on the job- -only in what you knowabout the performance of the tasks on the list. Your know-ledge of these tasks could have come fromn your training,observation, or past performance. Try to respond inevery instance with a scale value. There are scale cardsattached to every booklet. Remove these cards and keel)them in front of you as you make your judgments.

* to make your task easier, fill in one column at a timne.Start with the first skill, (demonstrate) finger dexterity(FD) and fill in that entire column starting with task I andcontinue until you have finished the last task in the column.Before you start filling in the scale values, read the defi-nition for the first skill and keep the definion list in front

58

of you, so that you can refer to it whenever necessary.After you finish the first column, read the definition ofthe second skill and complete the second column. If youcomplete the form in this manner, you will not have toremember the definitions of all of the skills at one time.

*9 after you complete each section of the form, go right onto the next. Make sure that every box has a response init. Do not leave any blank boxes.

* fill in the information called for on the cover page. Noneof the information you provide will be reported by yourname. We request your name for administrative pur-poses only. The data collected will be grouped and treat-ed statistically. All information provided is used forresearch purposes only and is held strictly confidential.

* if you have any questions, feel free to ask at any time.

7. During the administration, pass among the respondents and makesure that the forms are being completed in a downward sequence onthe page and that each respondent is referring to the proper definitionfor the column being completed.

8. Maintain an informal and friendly atmosphere so that a rapport be-tween the administrator and the respondents is developed.

9. Make extra pencils available to the respondents, as needed.

10. Go over the forms when they are turned in to assure that there areno blanks, glaring errors, or obvious inconsistencies, If any aredetected, ask the respondent to review the form.

Personal Interview

A semistructured personal interview was conducted with a sam-ple of the respondents in order to

1. provide an opportunity to acquire retest data fora test-retest reliability determination.

2. obtain evaluations of the appropriateness of theperceptual/ psychomotor abilities included in thetaxonomy.

3. obtain respondent reaction to the two scaling tech-niques.

4. acquire opinions on methods for improving the da-ta collection instruments

The full interview is presented in Appendix B to this report.

59

Interview er Instructions

The administrator of the survey forms also served as the inter -viewer. All interviews were conducted in private and by scheduled ap-pointment, on the day following the completion of the forms. The in-structions provided to the interviewer are presented below.

Interviewer Instructions-

1. Conduct the interview in a quiet place, free from traffic and exces-sive interference.

2. Introduce yourself and Applied Psychological Services.

3. Explain that the interview was designed to obtain opinions about thequestionnaire forms and also to obtain additional information aboutthe amount and performance quality variability levels of the variousperceptual/ psychomotor abilities involved in the tasks performedin the individual's career field. State the following:

I have a short interview which will, in part, be based on theform you completed. Your answers will be held confidential.The data will be treated statistically and your name will notbe associated with it in any way.

4. Conduct the interview in a friendly and informal manner. This inter-view is not a test, but a fact finding and opinion searching activity.Do not be critical, approving, or disapproving. Listen carefully toall comments; be attentive and supportive at all times.

5. Complete the general information called for on the interview formbefore starting the interview.

6. Ask the questions as they are worded, since standardization acrossall respondents is sought. Do not omit any portion of a question, oradd a comment that is not included. Read the questions slowly andclearly and in the same order in which they appear in the form.Present the interviewee with the card of options, wherever indicatedon the interview form. Repeat a question that was misunderstood ormisinterpreted; do not rephrase or reword the question. If the re-spondent needs additional time Lo think of a response, allow it.

7. Provide the respondent with positive feedback whenever possible bynodding your head, or passing neutral comments such as "yes,""1okay, " or "I see.

8. If a response is incomplete or irrelevant, probe for a mnore accept-able response. This may be accomplished b 'N repeating the questionor the response. Other effective probes are to ask if there is "anly-thing else," "how so,"1 or "Ican you explain that to me."

60

9. Record responses in the spaces provided on the interview form at

the time the responses are provided.

10. It is important that the interviewee understands the polarity of therating scales. For example, zero means very unimportant or veryeasy.

11. Answer all questions in a neutral manner. Terminate the interviewby thanking the individual for cooperating.

Interview Sample

Twenty-four interviews were conducted--12 representing job in-,cumbents in each of the two career fields. In each career field, six su-pervisors and six subordinates were interviewed. Every supervisor, inboth career fields, had served in that career field and in the Air Forcefor over five years. The subordinates who were interviewed were some-what less experienced than their supervisor counterparts. As a group,they were equally divided in their career field experience. Approximate-ly one-half had served three to five years and the other half had servedmore than five years in their career field.

Results--Quantitative

A set of data analyses was completed in order to determine theutility of the methods and taxonomy. These analyses sought

1. To establish the ability of the 13 taxonomic classesto differentiate among tasks.

2. To examine the relationship between the supervisoryand the subordinate perceptions of job perceptual/psy-chomotor skill influences.

3. To examine the relationship between the amount ofthe abilities required in the performance of the tasksand the performance quality variability.

4. To develop profiles of perceptual/ psychomotor abil-

ities for each specialty.

5. To determine the relationship between the categori-cal and the magnitude estimation judgments and thenature of the underlying scale of each.

6. To evaluate interrater agreements.

7. Tc obtain measures of test-retest reliability.

G 1

Discrimination- -Taxonomic Classes--Fire Protection Career Field

To obtain information about the ability of the taxonomic classesto discriminate, measures of central tendency were calculated for the13 perceptual/ psycho motor abilities. For the Fire Protection (57 lXX)career field, these were calculated separately for the supervisory rat-ers and subordinate raters, for both scaling techniques, as well as forthe amount and performance quality variability judgments. The resultsare presented in Table 3-3.

The mean data for the categorical scale evaluations of the FireProtection career field indicated, at best, only a moderate range acrossthe 13 taxonomic classes. The range of the mean value for the supervi-sors was 1. 9 to 3. 4, while the range for the subordinates was 2. 4 to 3. 3.The inidpoint of these values was approximately 2. 9, in both cases.These ranges seem somewhat restricted. The standard deviations a-round the means showed reasonable spread for each taxonomic class.The standard deviations ranged from 1. 2 to 1. 6 and 1. 4 to 1. 7, for thesupervisors and subordinates, respectively.

For the magnitude estimation method in the Fire Protection ca-reer field (Table 3-3), the range of mean values for the 13 abilities was28 to 53 for the amount judgments of the supervisors (with a spread ofstandard deviations ranging from 25 to 38) and from 38 to 55 (with stand-ard deviations ranging from 26 to 32) for the judgments by subordinates.Again, these ranges seem somewhat restricted. As with the categoricalscale, the approximate midpoint of the evaluations was at the low endof the "imoderate"i portion of the magnitude estimation scale(43 for thesupervisor group and 46 for the subordinates).

The data for the performance quality variability indicated a less-er spread of mean data and standard deviations, as well as a consistent-ly lower set of scale value selections than the amount data. Table 3 -3 in-dicates a range of mean data on the categorical scale judgments, acrossthe 13 abilities, of 1. 7 to 2. 5 (with standard deviations ranging f rom 1 , 2to 1. 4) and 2. 2 to 2. 6 (with standard deviations ranging from 1. 3 to 1. 5)for the supervisors and subordinates, respectively. Similarly, the per-formance quality variability magnitude estimation scale judgments indi-cated less spread than the amount judgments and lower scale Choices,e. g., a range of 26 to 45 for the supervisors (with a standard deviationrange from 26 to 31) and 35 to 45 for the subordinates (with a standarddeviation range from 27 to 41).

While the mean scale values for amount and for performancequality variability were distributed in close proximity, making it

62

Table 3-3

Perceptual/ Psychomotor Ability Means andStandard Deviations for the Fire Protection Career Field

14 0

F Categorical Scale

P u Amount Performance Quality Variability) W Supervisor Subordinate Supervisor Subordinate

M a M o M a M a

FD 2.7 1.4 2.8 1.4 2.2 1.3 2.2 1.3MD 3.1 1.5 3.2 1.4 2.5 1.4 2.5 1.4CP 3.1 1.5 2.9 1.5 2.4 1.4 2.4 1.3RC 3.1 1.6 2.7 1.6 2.4 1.4 2.3 1.4RM 3.4 1.4 3.1 1.4 2.3 1.3 2.6 1.4VA 2.9 1.5 3.0 1.5 2.4 1.3 2.3 1.3PM 3.1 1.5 3.3 1.5 2.3 1.3 2.6 1.4AD 2.5 1.5 2.9 1.5 2.2 1.4 2.4 1.4AD 2.5 1.5 2.7 1.5 2.2 1.4 2.6 1.3CLP 1.9 1.2 2.4 1.5 1.7 1.2 2.5 1.5DP 2.7 1.6 2.9 1.7 2.2 1.4 2.5 1.4DA 2.6 1.5 2.7 1.5 2.1 1.3 2.4 1.3DM 2.3 1.4 2.4 1.5 2.0 1.3 2.5 1.4

-' 2.3 1. . . . . . .

-4 00 Magnitude Estimation Scale

0. 0JWX-H Amount Performance Quality VariabilityCjU -4$4 Supervisor Subordinate Supervisor SubordinatePA M a M a M M

FD 42.8 30.1 46.5 27.1 37.7 29.8 35.4 28.0MD 50.8 31.3 55.1 26.1 41.5 29.5 39.0 29.4CP 47.6 31.3 50.5 29.0 41.1 29.0 40.4 29.2RC 45.0 33.3 45.7 30.9 42.3 30.5 43.6 29.1VM 49.8 30.1 49.4 29.9 44.8 30.7 41.0 29.4VSA 49.8 36.9 48.6 31.9 39.6 30.1 40.7 29.7PM 52.8 35.7 52.3 31.3 45.1 30.4 43.8 29.2AD 41.7 38.1 46.8 30.6 35.3 29.2 42.3 28.9AM 42.3 31.1 44.1 29.5 33.9 29.9 42.1 40.9CLP 28.2 25.0 38.1 28.2 26.3 27.3 42.3 31.3DP 40.6 32.2 47.1 29.9 33.5 28.8 41.8 29.7DA 37.2 28.7 44.2 28.3 33.9 25.6 37.7 27.2

37.9 29.5 44.7 30.5 30.6 26.9 44.5 29.5

63

difficult to separate the perceptual/psychomotor abilities into high andlow groups, an arrangement into high and low groups of abilities waspossible in each case. The mean scale values for the abilities werehierarchically arranged (from high to low) and lines of demarcationwere drawn between the largest gaps, separating the abilities into highand low groups (on each scale). The results are as follows:

Amount

Magnitude

Categorical Scale Estimation Scale

High:

Supervisors Manual Dexterity Manual DexterityControl Precision Visual MemoryRate Control Visual Speed andVisual Memory AccuracyPosition Memory Position Memory

Subordinates Manual Dexterity Manual DexterityVisual Memory Control PrecisionVisual Speed and Position Memory

AccuracyPosition Memory

Low:

Supervisors Clerical Perception Clerical PerceptionKinesthetic Memory

Subordinates Clerical Perception Clerical PerceptionKinesthetic Memory

Performance Quality Variability

Magnitude


High:

Supervisors Manual Dexterity Visual MemoryControl Precision Position MemoryRate ControlVisual Speed and

Accuracy

Subordinates Visual Memory Rate ControlPosition Memory Position MemoryAuditory Memory Kinesthetic Memory

64

Low:

Supervisors Clerical Perception Clerical Perception

Subordinates Finger Dexterity Finger Dexterity

This suggests that, in spite of the somewhat restricted rangeyielded by the taxonomy and scales, some separation can be achieved.Moreover, the "high" and the "low" results seem to make sense interms of what is generally known about the duties of the Fire Protectioncareer field.

Discrimination-Taxonomic Classes--Munitions Maintenance Career Field

['able 3-4 presents the 13 taxonomic class means and standarddeviations for both the categorical and magnitude estimation scale judg-ments for the Munitions Maintenance career field. The range for the su-pervisors on the categorical scale, for the amount judgments, was from1. 4 to 2. 9 (with a range of standard deviations from 0. 9 to 1. 5) and therange for the subordinates was from 1. 3 to 2. 9 (with a range of standarddeviations from 0. 8 to 1.6). On the magnitude estimation scale, the rangeof means for the supervisors was from 15 to 40 and it was from 24 to 48for the subordinates. rhe standard deviations, for the respective groupsranged from 15 to 30 and frorn 22 to 57. Again, the somewhat restrictedrange of the responses is evident.

For quality variability and the categorical scale, the range ofmean judgments was from 1. 6 to 2. 1 for the supervisors and from 1.7 to2. 1 for the subordinates. The standard deviations were again reasona-ble and ranged from 0. 9 to 1. 2 for the supervisors and from 1. 2 to 1. 5for the subordinates. The magnitude estimation scale produced the fol-lowing range of mean data for quality variability: 22 to 30 for the super-visors and 18 to 31 for the subordinates. The range of standard devia-tions for the supervisors was from 19 to 25 for the supervisors and it wasfrom 19 to 28 for the subordinates.

The distribution of mean judgments over the 13 abilities wasagain such that lines of demarcation could be readily interposed on a dis-tribution of the values, forming meaningful relative groups of "high" and"low" abilities. Such a result, at least partially, Supports the discrimi-nating power of the methods employed.

65

Table 3-4

Ability Means and Standard Deviations for the

Munitions Maintenance Career Field

r-4 0

: 0 o Categorical Scale

0. 0 4j

Amount Performance Quality Variability

0 c . Supervisor Subordinate Supervisor Subordinate

9w 20 M a H .4 1. a 1 1

FD 2.0 1.1 2.3 1.4 1.8 1.2 1.9 1.3ND 2.3 1.4 2.5 1.5 1.8 1.1 1.9 1.3CP 2.1 1.4 2.0 1.4 1.8 1.2 1.8 1.2RC 1.8 1.3 2.0 1.5 1.7 1.1 1.7 1.2

VM 2.9 1.5 2.9 1.5 2.1 1.2 2.1 1.3VSA 2.6 1.5 2.7 1.6 2.1 1.2 2.0 1.3PM 2.6 1.5 2.7 1.5 1.9 1.2 2.1 1.4

AD 2.0 1.4 2.2 1.5 1.7 1.2 1.9 1.4AM 1.8 1.2 1.9 1.4 1.6 1.2 1.9 1.5CLP 2.3 1.5 2.6 1.5 1.8 1.1 2.0 1.4DP 1.9 1.3 2.2 1.5 1.7 1.1 1.9 1.4DA 1.8 1.2 2.2 1.4 1.6 0.9 1.8 1.2M 1.4 0.9 1.3 0.8 1.6 1.1 1.8 1.4

.-4 00 " Magnitude Estimation Scale

&. 0 >10) -V4 Amount Performance Quality VariabilityQ U r-4$>," Supervisor Subordinate Supervisor Subordinate

P4N a H a 14 a 14

FD 25.4 20.8 30.7 23.7 22.3 20.8 26.5 23.5

MD 33.8 27.7 38.2 28.1 24.1 21.0 30.5 27.4CP 31.2 27.6 29.9 27.4 25.0 22.3 26.9 24.8

RC 23.9 25.3 29.3 28.0 23.3 22.7 26.2 24.2

VM 39.7 28.4 45.6 30.0 29.6 23.0 30.9 25.4VSA 37.5 29.7 46.1 56.7 28.5 22.6 30.4 27.0PM 38.8 29.1 47.6 29.6 26.9 23.0 30.9 26.3AD 26.9 27.1 33.9 30.0 24.9 22.4 26.2 23.5

26.2 30.4 28.8 23.1 22.0 30.4 27.6. 27.j 40.3 30.5 24.3 21.5 27.0 23.2

J. 4 39.4 30.7 22.6 21.1 24.7 24.1

11'.3 29.5 24.0 24.6 23.5 21.521,5 21.5 19.2 18.2 19.4

AmountMagnitude


High:

Supervisors Visual Memory Visual MemoryVisual Speed and Visual Speed and

Accuracy AccuracyPosition Memory Position Memory

Subordinates Manual Dexterity Visual MemoryVisual Memory Visual Speed andVisual Speed and Accuracy

Accuracy Position MemoryPosition MemoryClerical Perception

Low:

Supervisors Kinesthetic Memory Kine sthetic Memory

Subordinates Kinesthetic Memory Kinesthetic Memory

Performance Quality VariabilityMagnitude


High:

Supervisors Visual Memory Visual MemoryVisual Speed and Visual Speed andAccuracy Accuracy

Subordinates Visual Memory Manual DexterityVisual Speed and Visual MemoryAccuracy Position Memory

Position MemoryClerical Perception

Low:

Supervisors Auditory Memory Finger DexterityDivided Attention Kinesthetic MemoryKinesthetic Memory

Subordinates Rate Control Kinesthetic Memory

67

Discrimination- -Tasks - -Fire Prote ction

In a similar vein, the means and standard deviations for the 20tasks (for amount and performance quality variability) on each scalewere calculated for the Fire Protection specialty. These appear inTable 3-5. The two sets of mean data, supervisor and subordinate re-spectively, for amount on the categorical scale ranged from 1. 7 to 3. 8and from 2. 0 to 3. 9. This result seems more acceptable than that in-dicated for the taxonomic classes. The standard deviations for thesemean data was entirely acceptable--l.1 to 1.6 for the supervisor groupand from 1. 3 to 1. 6 for the subordinate group. The magnitude estima-tion scale data indicated similar results for the amount date (Table 3-5).The range of mean data for the tasks, for amount, ranged from 24 to 63for the supervisors (with a standard deviation range of 21 to 52); the rangefor the subordinates was from 30 to 69 (with a standard deviation rangeof 24 to 31).

For the performance quality variability data (Table 3-5), the su-pervisors' range of mean amount judgments on the categorical scale was1. 4 to 3. 1; the range for the subordinates was 1. 8 to 3. 7. These rangesare again broader than for the parallel analysis in which taxonomicclasses were considered across tasks. The respective ranges of stand-ard deviations were 0. 8 to 1. 5 and 1. 2 to 1. 4. On the magnitude estima-tion scale, the range of performance quality variability means for super-visors across the 20 tasks, was 23 to 57 (with a standard deviation rangefrom 21 to 31). The range for subordinates, on the same scale, was 32to 65 (with a standard deviation range from 25 to 48).

Accordingly, the restricted range patterns seen for the individualtaxonomic classes was somewhat broken for the task data.

Discrimination - -Tasks - -Munitions Maintenance

Table 3-6 presents the means and standard deviations for theamount and performance quality variability judgments, on both scales,for each group of Munitions Maintenance judges. The range of the amiounitjudgments, for supervisors and subordinates, on the categorical scale,was 1. 7to 3. 1 and 1. 8 to 3. 2, respectively. These ranges seem quiteacceptable and are greater than the corresponding ranges for the taxono-ic class data. The range of standard deviations for the supervisors was1. 0to 1. 7 and 1. 1 to 1. 7 for the subordinates. The range on the magni-tude estimation scale for supervisors was 19 to 53 (with a standard devia-tion range from 16 to 33); the range for subordinates was 25 to 54 (with arange of standard deviations from 22 to 71). These values, too. are con-siderably greater than for the taxonomic class data. The mean categorical

68

Table 3-5

Task Means and Standard Deviations AcrossAbilities for the Fire Protection Career Field

Categorical Scale

Amount Performance Quality VariabilitySupervisor Subordinate Supervisor Subordinate

Task H a H a H a H a

1 3.8 1.4 3.9 1.4 3.1 1.5 3.7 1.42 2.5 1.4 2.6 1.5 2.0 1.1 2.5 1.43 1.7 1.1 2.0 1.3 1.4 0.8 2.0 1.34 3.4 1.4 3.3 1.5 2.3 1.2 2.5 1.45 2.7 1.5 2.7 1.5 2.2 1.2 2.4 1.46 2.6 1.4 2.6 1.4 2.0 1.2 2.1 1.27 3.4 1.5 3.2 1.5 2.6 1.4 2.8 1.48 3.3 1.6 2.9 1.5 2.7 1.5 2.4 1.49 3.1 1.5 3.0 1.5 2.7 1.5 2.6 1.4

10 3.3 1.5 3.0 1.4 2.4 1.4 2.5 1.211 3.1 1.5 3.1 1.4 2.4 1.4 2.6 1.312 3.4 1.4 3.0 1.4 2.5 1.4 2.4 1.313 2.3 1.4 2.4 1.5 1.8 1.2 2.0 1.314 2.8 1.5 3.0 1.6 2.5 1.3 2.5 1.3i5 2.2 1.3 2.5 1.4 1.7 1.0 2.0 1.216 2.0 1.2 2.3 1.4 1.4 0.8 2.1 1.317 3.4 1.5 3.3 1.4 2.6 1.4 2.8 LAt18 2.7 1.5 3.0 1.5 2.2 1.4 2.4 1.319 1.9 1.3 2.5 1.5 1.8 1.1 1.8 1.220 2.4 1.5 3.0 1.6 2.0 1.5 2.4 1.4

Magnitude Estimation Scale

Amount Performance Qulityf VariabilitySupervisor Subordinate Supervisor Subordinate

Task H a H a H 0 H a

1 63.1 30.6 68.9 29.0 56.8 31.1 64.6 30.42 34.5 27.7 43.9 30.1 32.2 28.1 39.7 29.43 25.2 21.8 30.1 23.5 22.8 20.8 37.7 30.44 46.2 30.1 48.8 31.4 43.0 30.5 42.0 29.95 42.5 29.8 42.9 29.2 39.8 29.2 38.3 28.16 39.8 28.7 44.2 27.3 32.8 28.0 31.9 25.07 52.9 31.4 50.8 28.8 46.8 30.7 48.5 28.68 49.3 31.6 47.2 29.8 43.7 29.7 40.1 28.59 48.3 30.3 50.5 29.4 42.7 29.2 44.1 26.9

10 51.4 31.0 51.7 29.7 39.7 28.8 39.3 25.911 51.3 31.0 53.7 29.3 39.7 29.3 42.6 26.212 52.0 51.7 50.0 28.6 39.4 29.9 39.7 28.013 35.3 30.0 38.6 30.1 31.2 27.7 38.5 47.814 48.5 31.0 50.7 29.9 44.1 29.8 39.0 25.715 30.6 25.1 39.5 28.3 28.6 25.1 33.2 26.716 23.7 21.4 34.0 25.1 25.8 24.0 36.1 28.817 54.7 31.9 55.3 29.2 45.3 30.1 46.0 28.518 44.9 30.7 49.7 29.1 36.8 29.7 36.3 27.7

19 34.6 29.2 44.2 2b.5 30.6 26.4 32.4 25.6

20 44.6 32.8 47.8 29.3 33.3 29.5 40.6 30.5

69

Table 3-6

Task Means and Standard Deviations forMunitions Maintenance Career Field

Categorical Scale


Supervisor Subordinate Supervisor Subordinate

Task M a H H a H a

1 2.4 1.5 2.5 1.5 2.0 1.3 2.0 1.32 2.0 1.3 2.4 1.5 1.6 1.0 2.1 1.5

3 2.3 1.5 2.3 1.5 1.8 1.1 2.1 1.4

4 1.7 1.1 1.8 1.2 1.5 0.9 1.6 1.2

5 1.8 1.2 2.0 1.4 1.7 1.1 1.5 0.9

6 1.7 1.2 1.8 1.3 1.4 0.8 1.3 0.8

7 2.2 1.4 2.3 1.4 1.7 1.0 1.8 1.2

8 1.8 1.0 2.0 1.3 1.4 0.8 1.5 1.19 2.2 1.4 2.3 1.4 1.7 1.0 2.0 1.3

10 2.8 1.5 3.0 1.6 2.4 1.3 2.5 1.5

11 3.1 1.6 3.2 1.6 2.6 1.4 2.7 1.5

12 1.7 1.0 1.8 1.1 1.4 0.8 1.5 1.0

13 2.4 1.4 2.6 1.4 2.1 1.3 2.1 1.4

14 2.2 1.4 2.4 1.5 1.9 1.2 1.9 1.3

15 2.1 1.3 2.0 1.3 1.7 1.0 1.7 1.2

16 2.6 1.7 2.7 1.7 2.1 1.4 2.2 1.5

17 1.9 1.3 2.3 1.5 1.7 1.1 1.9 1.4

18 1.9 1.3 1.9 1.4 1.6 1.0 1.6 1.1

19 2.2 1.3 2.2 1.3 1.7 1.1 1.8 1.2

20 2.1 1.3 2.5 1.5 1.6 1.3 2.1 1.5

Magnitude Estimation Scale



Task H a M a M a M a

1 34.8 29.7 "9.3 29.7 26.0 22.6 26.6 23.5

2 27.4 26.5 40.2 71.4 22.5 19.7 25.2 23.0

3 31.4 26.9 35.5 29.5 23.1 20.9 26.6 23.5

4 20.3 18.7 24.5 22.3 19.5 17.9 18.9 18.6

5 23.8 23.3 30.6 27.6 21.4 19.2 23.3 22.3

6 22.0 22.0 28.6 26.0 18.1 1..9 22.5 20.9

7 30.1 25.3 33.7 25.8 23.1 18 2 27.0 23.2

8 19.8 17.7 26.0 22.7 17.4 15.9 20.0 19.1

9 31.6 26.3 36.6 28.4 24.0 19.7 24.9 23.1

to 46.8 30.8 52.1 31.5 39.5 27.3 39.5 29.1

11 52.5 31.9 54.4 31.5 43.0 28.0 42.8 29.5

12 18.8 15.5 26.3 22.2 18.0 16.1 19.4 18.5

13 35.6 27.2 41.7 27.6 29.8 23.3 32.4 26.1

14 30.2 25.8 40.8 30.7 27.3 24.0 25.9 23.6

15 30.8 27.2 33.4 26.5 22.1 19.1 23.1 20.7

16 40.3 33.4 43.2 32.6 29.3 26.7 32.7 29.0

17 26.1 25.4 36.1 29.3 23.8 22.% 28.1 25.6

18 25.1 26.5 31.7 29.0 20.5 18.b 20.6 19.6

19 28.6 24.4 37.5 27.9 21.6 18.3 26.7 24.0

20 25.7 23.4 41.4 30.4 18.7 17.8 35.4 29.6

70

scale performance quality variability data for the supervisors achieveda range from 1.4 to 2.6 (with a standard deviation range from 0.8 to 1.4).The range for the subordinates was 1.3 to 2.7 (with a standard deviationrange of 0. 8 to 1. 5). The magnitude estimation scale produced super-visor mean range of 17 to 43 (with a standard deviation range of 16 to 28)and a subordinate mean range of 19 to 43 (with a standard deviationrange of 19 to 30).

Discussion and Summary of Discrimination Analyses

The prior sets of analyses suggest that the techniques do littleto discriminate among taxonomic classes when abilities are judgedacross tasks. On the other hand, the techniques seem more sensitivewhen tasks were judged across individual taxonomic classes. Table 3-7summarizes the range of mean values yeilded by the separate methodsand rater levels by career fields. In all corresponding cells. the rangefor task ratings across taxonomic classes is greater than the rangefor taxonomic classes across tasks. The ranges for the task ratingsacross classes seem to suggest acceptable sensitivity but the same cannot be said for the taxonomic class across task data. This suggests thatdata of the nature here involved are best sought at the task level.

Regardless of scaling approach or set of judges, the data suggestthat the two career fields involved in the pretest are not heavily loadedin perceptual/ psychomotor requirements of the type included in our tax-onomy. All distributions were skewed to the left. While the possibilityof rater error exists, such an explanation does not seem tenable in viewof the controls instituted, the diver,.ityof the two career fields involved,and the experience range of the two subject groups.

Amount and Performance Quality Variability Comparisons

A visual inspection of the mean data presented in 'Fables 3-3 and3-4 suggests a degree of association between the two judged variables- -

amount and performance quality variability. In order to examine furtherthe association between these variables, product moment correlation co-efficients were calculated. The resultant correlation coefficients arepresented in Table 3-8.

'Fhe correlation between the judgments of amount and perform-ance quality variability was moderate to high in six of the eight compar-isons. A comparison of the supervisor and the subordinate cor-relationcoefficients in Table 3-8 indicates, in all cases, slightly depressed co-efficients for the subordinates. This suggests that the supervisors tend-ed to judge a task as having greater performance quality variability for

71

of

4~ -

-44 -4 4

C,4 -T

041-44

Cd

0%Coc

Co

w 4

~E 0)L

0 0O 61(

41 ) (Ui ,-4 4Cm

Cd C2 0i

0- 0(to ca) 0 ) a .

a 4 0)q

-0 Q)CI $4 JJi

to 41 -4 (

0)0 U-*0

Cd0 £1: ) U $

00

0) 0

Cd (U0 )

00

.c: Ol

0004U 1 4i

U (0

72

Table 3-8

Product Moment Correlation Between Amountand Performance Quality Variability (N = 13 Abilities)


Variable r

Supervisor, categorical scale .87Subordinate, categorical scale .23Supervisor, magnitude scale .93

Subordinate, magnitude scale -.15

Meanr .65


Supervisor, categorical scale .90

Subordinate, categorical scale .76

Supervisor, magnitude scale .84

Subordinate, magnitude scale .65

Meanr = .81

Grand Meanr = .74

73

a particular ability when the task required a greater amount of the abil-ity for successful performance. Conversely, those tasks that variedless in the performance quality variability, as a function of a specificability, were regarded by the supervisors as requiring less of the abili-ty for successful performance. Subordinates, on the other hand, seemedto make slightly more independent judgments on the amount and perform-ance quality variability factors.

Moderately high correlations were obtained when the averagecorrelation (obtained by averaging the z-coefficient equivalents) was ob-tained for each specialty--. 65 and . 81 for Fire Protection and MunitionsMaintenance career fields, respectively. The estimate of the populationvalue was. 74 when the eight separate correlation coefficients were aver-aged.

The correlational data, taken as a group, suggest that the twoquestions are moderately interdependent. There is an indication thatsome association exists between the perceptions of the amount of anability required to perform a task and the performance quality variabil-ity it produces. However, from the point-of-view of job analytic thor-oughness, it seems that both questions may best be considered in anyjob analysis which aims to be complete.

Supervisor and Subordinate Comparisons

If the judgments of supervisors and subordinates can be demon-strated to be associated, then a position could be taken that supervisor-subordinate judgments may be combined prior to the analytic treatmentof the data, and in future work, distinctions between the two groupsmay be disregarded. In this regard, note the similarity between themean data of the supervisors and the subordinates for tasks and forperformance quality variabilit reported in Tables 3-3, 3-4, 3-5, and3-6.

Table 3-9 presents the correlation between supervisor and sub-ordinate judgments for amount and performance quality variability sep-arately, by rating scale type. The correlation coefficients ofTable3-9show a close association between the supervisor and subordinate rat-ings of amount, regardless of type of rating scale.

The average correlation between supervisory and subordinatejudgments in the Fire Protection specialty, across rating scales,amount, and performance quality variability was .54. An average of.84 was obtained for the Munitions Maintenance s)ecialty. An esti-mate of the population value was obtained by averaging the correlationsacross both specialties. The obtained correlation, .72, was consider-ed moderately high.

74

lawi

Table 3-9

Product Moment Correlations BetweenSupervisors and Subordinates (N = 13 Abilities)


Variable r

Amount, categorical scale .79

Performance quality variability, categorical scale -.10

Amount, magnitude scale .91Performance quality variability, magnitude scale -.09

Meanr .54


Amount, categorical scale .94

Performance quality variability, categorical scale .63

Amount, magnitude scale .91Performance quality variability, magnitude scale .67

Meanr = .84

Grand Meanr .72

75

Accordingly, some basis exists for combining the data producedby the two sets of respondents or for using only one or the other groupof respondents in job analytic studies of the type considered here.

Test-Retest Reliability

in order to collect data relative to the response stability withinjudges, the 24 raters who participated in the personal interview reeval-uated a sample of 10 tasks on each of the 13 perceptual/psychomotortaxonomic classes. The group was divided such that half of each of thefour groups of raters (supervisors-Munitions Maintenance, subordi-nates-Munitions Maintenance, supervisors -Fire Protection, subordi-nates-Fire Protection) reevaluated 10 tasks using either the categoricalor the magnitude estimation scale.

The Lawlis and Lu (1972) approach was used to determine theagreement of ratings on two separate occasions. An advantage of thistechnique over other conventional reliability measurements (e. g., cor-relation) is in its power to measure relationships when the total vari-ance of range is small, as is often the case with rating scales and inthe case of the present data set. Within the Lawlis and Lu approachthe null hypothesis of agreement by chance is tested through the chi-square formula (with one degree of freedom):

X2 = (N 1 - Np - .5) 2 + (N 2 - N(l - p) -. 5)2

Np N ( - p)

where: N = number of abilities being ratedN1 = number of observed agreementsN2 = number of observed disagreementsp = the probability of k judges achieving

agreement by chance.5 = correction for continuity

A statistically significant X2 value indicates that the observedagreement is not due to chance. The p value in the formula is obtain-ed from a table of probabilities of chance agreement, based on a rec-tangular distribution model (i.e., every judgment has the same proba-bility of occurring under the hypothesis that the judges have no under-standing of the scale and their ratings are random). The determina-tions of agreement were based on the most rigorous criterion--identi-cal ratings on the two occasions. This is an especially stringent cri-terion. The resultant X2 values are presented in Table 3-10.

The coefficients of agreement resulting from this analysis werehigh and indicate non-chance agreement over the two occasions.

76

Table 3-10

Coefficients of Agreement BetweenTwo Evaluations of the Same Tasks

Rater Chi-Square


Supervisor, amount-categorical 17.97*

Supervisor, amount-magnitude 24.31*

Subordinate, amount-categorical 24.31*

Subordinate, amount-magnitude 17.97*

Supervisor, performance quality variability-categorical 24.31*

Supervisor, performance quality variability-magnitude 12.50*

Subordinate, performance quality variability-categorical 24.31*

Subordinate, performance quality variability-magnitude 49.12*


Supervisor, amount-categorical 12.58*

Supervisor, amount-magnitude 8.16*

Subordinate, amount-categorical 2.20

Subordinate, amount-magnitude 49.12*

Supervisor, performance quality variability-categorical 49.12*

Supervisor, performance quality variability-magnitude 2.20

Subordinate, performance quality variability-categorical 17.97*

Subordinate, performance quality variability-magnitude 24.31*

Statistically significant at or below the .01 level of confidence.

77

-Rump

Interrater Agreement

Agreement between raters was calculated for each perceptual/psychomotor taxonornic class by rater type (supervisor and subordi-nate), type of scale (categorical and magnitude estimation), and evalu-ative factor (amount and performance quality variability).

To accomplish these interrater reliability estimates, tables wereconstructed which contained the number of responses for each scale val-ue selected by task. In this process, the magnitude estimation scale re-sponses were collapsed into five category units, such that responses be-tween 0 and 19 were counted as response category 1; responses between20 and 39 were counted as response category 2, etc.

To determine the percentage of agreement on each ability, thefollowing steps were taken:

1. For a given data set and each task within theset, select the rp~cponse category containingthe largest number of respon~ses as the mod-ulus.

2. For each task, sum the number of responsesin the modulus category with those in the tworesponse categories on the sides of the mod-ulus (if the modulus is the first or last re-sponse category, add all responses in the twounits closest to the modulus). This is theagreement sum.

3. Sum the agreement sums over all 20 tasks.

4. Obtain the total number of responses made forall 20 tasks (e.g. , if there were 24 raters andeach evaluated 20 tasks. then the total numberof possible responses is 480).

5. Divide the result from step 4 into the resultfrom step 3 to obtain the percentage of agree-ment among raters.

Tables 3-11 and 3-12 present the resultant interrater agreementpercentages for the categorical scale and the magnitude estimationscale, respectively. The average percentage of agreement was obtain-ed for each perceptual/ psychomotor ability, across all ratings in bothspecialties and separately by specialty, rater type, and evaluation fac-tor. The two tables indicate that the minimum percentage of agreementwas 76 on both the categorical and magnitude estimation scales for thePosition Memory ability and for Visual Memory on the categorical scale.

78

Table 3-11

Percentage Agreement AmongRaters for the Perceptual/Psychomotor Abilties

on the Categorical Scale Judgments

1-4 0

Fire Protection Munitions Maintenance

I 4w w Supervisor Subordinate Supervisor Subordinate

A PQV A PQV A PQV A PQV* Mean

FD 75 82 74 80 87 85 79 85 81

MD 79 78 79 76 84 89 82 83 81

CP 82 80 73 81 87 89 89 88 84

RC 80 79 79 80 92 90 90 86 85

VM 73 79 71 74 73 82 73 81 76

VSA 69 77 74 83 77 83 74 81 77

PM 74 76 72 78 71 87 72 78 76

AD 83 80 75 78 81 87 82 84 81

AM 78 75 70 76 84 88 74 80 78

CLP 84 86 71 76 74 87 73 81 79

DP 79 81 73 75 93 91 87 84 83

DA 82 85 76 80 89 93 79 87 84

KM 79 83 71 73 97 90 97 79 84

Mean 78 80 74 78 84 88 81 83

* A = Amount

PQV = Performance Quality Variability

79

7 AD-A093 981 APPLIED PSYCHOLOGICAL SERVICES INC WAN E PA F/ 5/10PERCEPTUAL/PSYCHONOTOfi REQUIREMENTS BASIC TO PERFORMANCE IN 35 --EYC(U)

2L DEC 80 A I SIEGEL, P J FEDERMAN, E H WELSAND F33615-78-C-0032UNCLASSIFIED AFHRL-TR-80-26 NL.urnuuIIuIuIuuIIIIIIIIIIIIII...fl-ElllllllEEEEEmuIIuuuIuIIIuu-IIIIIIIIIIIu-EIIIIIIEEIIE-mIIIIEEEIII

I.I

I I 1 11 -1 0Table 3-12

* Raters Percentage Agreement AmongSRaters for the Perceptual/Psychomotor Abilities

on the Magnitude Estimation Scale Judgments

Fire Protection Munitions Maintenance


A PQV A PQV A PQV A PQV* Mean

[ D' 69 77 77 81 92 93 89 91 84

"D 74 76 77 79 86 93 83 85 82

P 77 78 76 79 94 94 92 91 85

RC 77 81 83 82 97 93 93 93 87

VM 69 70 72 78 80 90 74 88 78

VSA 71 78 72 79 83 91 81 86 80

PM 68 70 67 77 79 89 70 85 76

AD 83 86 79 83 92 95 88 94 88

AM 85 87 84 80 93 93 89 91 88

CLP 95 94 88 86 85 93 80 91 89

DP 79 86 80 82 94 94 85 92 87

DA 83 91 83 87 89 91 85 95 88

KM 86 91 80 81 99 96 96 97 91

Mean 78 82 78 81 89 93 85 91

• A = Amount

PQV = Performance Quality Variability

80

The average percentage of agreement ranged from 76 to 85 for the cat-egorical scale judgments and from 76 to 91 for the magnitude estimationscale judgments.

The agreement among the raters, in all cases, seems accepta-bly high. It seems that the raters uniformly understood and interpretedthe rating methods and the definitions of the taxonomic categories. Sucha finding supports the usefulness of the methods and procedures employ-ed.

Intercorrelation Among Abilities

The taxonomic classes were selected with consideration of uni-dimensionality (uniqueness) as a partial criterion. However, some in-sight into the empirical relationship among the various abilities was

required.

The perceptual/ psychomotor ratings were intercorrelated to es-tablish the relationship among the abilities. The resultant intercorrela-tion matrices are presented as Appendix C to this report.

Frequency distributions (in which supervisor and subordinatedata were combined) of the results are presented as Tables 3-13 and3-14. The frequency distributions were constructed to reflect the log-ic that the class interval < . 30 = little or no relationship, the class in-terval . 31 to . 60 = moderate relationship, and > . 61 = high relationship.Most of the intercorrelations in both tables are in the below .30 or the.31 to .60 ranges. This indicates some degree of independence or u-niqueness among the perceptual/psychomotor abilities.

The perceptual/psychomotor taxonomic classes that were high-ly intercorrelated with others were examined to determine the elementsthat were common to the intercorrelated abilities.

Four different correlations of . 80 or above occurred with somefrequency in both the amount and performance quality variability judg-ments. These involved the following taxonomic classes: (a) ManualDexterity and Control Precision, (b) Visual Memory and Position Mem-ory, (c) Visual Speed and Accuracy and Position Memory, and (d) Au-

ditory Discrimination and Auditory Memory. The common elements inthe Manual Dexterity and Control Precision abilities are the arm andhand manipulations and movements. The major difference between themis that in one case objects (such as tools and materials) are manipulated,whereas control mechanisms (such as levers and pedals) are manipulat-ed in the other. The obvious common element in the Visual Memory and

81

Table 3-13

Frequency Distribution of Intercorrelations AmongTaxonomic Classes for Amount

Amount-Categorical Scale

Fire Protection

Correlation N Percent

< .30 26 17.31- .60 94 60> .61 36 23

156

Munitions Maintenance

< .30 95 61.31- .60 47 30> .61 14 9

156

Amount-Magnitude Estimation Scale

Fire Protection


< .30 16 10.31 - .60 65 42

> .61 75 48156


< .30 100 64.31 - .60 42 27

> .61 14 9156

82

LI!

Table 3-14

Frequency Distribution of Intercorrelations AmongTaxonomic Classes for Performance Quality Variability

Performance Quality Variability-Categorical Scale

Fire Protection


< .30 20 13.31 - .60 79 51> .61 57 37

m 156


< .30 8 5.31 - .60 42 27> .61 106 68

156

Performance Quality Variability-Magnitude Estimation Scale

Fire Prote tion


< .30 8 5.31 - .60 43 28

> .61 105 67156


< .30 28 18.31 - .60 80 51> .61 48 31

156

83

the Position Memory classes is the reliance on memory, to recall thingsthat were seen in the past for- the first class and to recall the positionof things from past experience in the latter class. The reason for theassociation between the Visual Speed and Accuracy class and the Posi-tion Memory class is not entirely clear. Finally, Auditory Discrimina-tion and Auditory Memory share the common element of sound discrim-ination.

Categorical and Magnitude Estimation Scale Comparisons

The data presented in prior sections suggested somewhat closeagreement in the results produced by the two different scaling ap-proaches employed in the pretest. Several analyses were performed todetermine the statistical relationship between the two different ratingscale approaches.

Table 3-15 presents the product moment correlation between thecategorical and the magnitude estimation scaling for various situations.The correlation coefficients were considerably higher for the MunitionsMaintenance raters as opposed to the Fire Protection raters. Withinboth specialties, the supervisors tended to be more consistent acrossthe rating scales than their subordinates. However, averaging the cor-relations across both specialties to obtain an estimate of thle populationvalue resulted in a relatively high degree of relationship between thetwo scaling approaches, r =. 80.

The relationship between the two scaling procedures was furtherexamined visually. Graphs of the mean results from the two scaling ap-proaches are presented in Figures 3 -1 through 3 -4. To plot the two setsof results on the same axes, a set of equivalencies of scale intervalmidpoints was established. The magnitude estimation scale was divid-ed into five segments: 0 to 19, 20 to 39, 40 to 59, 60 to 79, and 80 to100, and the midpoint of each magnitude estimation scale segment wasequated to each scale point on the categorical scale in the followingmanner:

Magnitude Midpoint Categorical Scale Value

10 130 250 370 490 5

The supervisor and subordinate judgments were combined.

84

Table 3-15

Product Moment Correlation Between Categorical and

Magnitude Estimation Scaling Approaches (N = 13 Abilities)

Fire Protection

Variable r

Supervisors, amount .53

Supervisors, performance quality variability .61

Subordinates, amount .90

Subordinates, performance quality variability .52

Meanr .69


Supervisors, amount .96

Supervisors, performance quality variability .86

Subordinates, amount .91

Subordinates, performance quality variability .64

Meanr = .88

Grand Meanr - .80

85

~40

30

-MAGNITUDE ESTIMATION SCALE

------ ]wCATEGORICAL SCALEI

Fo mo CP RC Um VS4 Pm AD A L O A g

PERCEPnhAL/PSYCHOMO0TOR ABIUTrY

Fiqxq 3-I. Corrvepod ,c bolwbef two scaling aproacties for the Fire

P otect ion spea~ty on l om.i

86

40

qjj v

I - MAGNITUDE ESTIMATION SCALE

-CATEGORICAL SCALE

11 UD 11 1 - .. t . I. IA

PERCIEPWhAL/PSYCHON" ABILITY

Fiqw 3-2. Cowvespondas buewu mom ?w Wng app -ad es far"* Fire

87

t I'I \

, '

A z!

I- , I/z I\ I I

\I I

II \' I\ I

I \ I 2

\1 ~

-iMAGNITUDE ESTIMATION SCALE20 CATEGORICAL SCALE

FD m cp PC vm VSA Pm D AM CLP DP A KM

PERCEPTUAL/ PSYCHOMOTOR ABILITY

Ftgre 3-3. Cmspoflndence behoen I swlin oap1odeforMVntiorn

Montsflanc. sCIOlty on amwoit.

88

-MAGNITUDE ESTIMATIONI SCALE

CATEGORICAL SCALE

PU 140 CP PtC vm VSA PM AU AM CLIN Ob a& KM

PERCEPTUAL/ PSMWOMR ABIITY

Fiue3-4. Coryswiwme bome two ncakno apmood.e for I~autwo~gmmietsa

89

The graphs all show close similarities in shape and elevationwith the exception of Auditory Memory in Figure 3-3.

Both the correlational data and the graphic presentations sug-gest that the task-taxonomic category judgments are not method sensi-tive. This holds for both the amount and the performance quality vari-ability factors.

Properties of Scales

Another set of insights into the nature of the scales provided bythe two scaling approaches is provided by plotting the means in ascend-ing order and showing the standard deviation of each mean on the sameplot. If the standard deviations increase at the scale extremes, thereis a suggestion that the raters are more sensitive to the scale extremesthan to the central area. Such a scale property is not usual in psycho-logical continua but is usual for physical continua. Quite obviously scalesof the latter type (prothetic scales) do not possess a constant Weberfunction and are less preferred than scales of the former type (meta-thetic scales).

The data in Tables 3-3 and 3-4 are presented graphically in Fig-ures 3-5 and 3-6. As a result of the agreement between the supervisorand subordinate judgments, as indicated by the prior analyses, the judg-ments of the two groups of judges were combined. The graphs presentthe 13 ability means, and their standard deviations, in an ascendingscale value order. Separate graphs are presented for the amount andperformance quality variability, on each of the two types of rating scales,for each of the two specialties pretested.

In addition, the graphs of the mean data place the 13 abilities ina perspective relationship to each other. For example, Figure 3-5 in-dicates several abilities that are high on the amount of the abilities re-quired to perform the listed tasks (e.g., Manual Dexterity, Visual Mem-ory, Position Memory) and several that are low on the amount of theabilities required (e. g., Clerical Perception and Kinesthetic Memory).The plots indicate almost constant standard deviations regardless oftype of scale (categorical or magnitude) or factor judged (amount or per-formance quality variation). This suggests the utility of either type of

scale for job analytic studies of the present type.

Combined Profiles

Figures 3-5 and 3-6 presented separate profiles of the amount

and the performance quality variability involvement of the abilities

90

4WIN

am

aD4 a n a • o

2.4 • •122 1.2a

AMaOUNT - A

OUWTY VANfIUTY -?AIT.GGt.Ak

T . . .M ,. N-. . . . •

at F m J$ am,

Sa . . . .l . . . . . a

FD = Finger Dexterity PM = Position MemoryAD = Auditory Discrimination MD = Manual DexterityCP = Control Precision AM = Auditory MemoryRC = Rate Control DP = Depth PerceptionVM = Visual Memory DA = Divided Attention

VSA = Visual Speed and Accuracy KM = Kinesthetic MemoryCLP = Clerical Perception

Figure 3-5. Means and standard deviations of taxonomicclasses for Fire Protection career field.

91

No. c, AO 390.

AV~K Ct A "OFOMQ Er AVT a:

IsO. AF

AMOLOW-ATIODWA OUALIW VARIABUr( -CAIEGWCA

so 10 * aU

so .i W

ANO--MGMl UvAT*

2P a Coto Prciio AMM =MJ Auior emrRC ~ ~ ~ ~ O =P Rat Control DP = etereto

CL = Clria 2erceptio

Fiue36.Masad tnaddeitos ftxnmi lse

forg~ Munitionsa.im Maitenncecarer

IS a 9a

a *~ '--No

represented by each taxonomic class in the two technical specialties.Another approach to portraying the taxonomic class involvement is tocombine the amount and the performance quality variability data into oneoverall index. The combinatorial approach selected was the geometricmean:

Total Involvement = v(A 2 + PQV 2 )

where: A = amount

PQV = performance quality variability

This is the familiar "city block" model which considers the twovectors to be orthogonal. Of course, other combinatorial techniquesare possible and possibly more defensible.

The combined profiles are presented as Figures 3-7 and 3-8.These represent the overall involvement of the taxonomic classes ineach of the two Air Force career fields included in the pretest.

Interview Findings

A semistructured interview was conducted with 24 of the super-visors and subordinates who participated in the major, pretest data col-lection effort. The interview attempted to inquire into problems associ-ated with the data collection techniques, the alternate scaling approaches,the taxonomy, and the like from the point of view of the respondent.

Two abilities were suggested in response to the question "Whatother perceptual/psychomotor abilities are required in the performanceof tasks that were not contained in the data collection forms?" Most re-spondents thought that the taxonomy was complete and thorough. Severalinterviewees, in both career fields, suggested that strength and physicalstamina 2 were important, e.g., for the continued handling of heavy mu-nitions in the Munitions Maintenance career field and in the rescue ofpersonnel (described as an activity that "exhausts and drains peoplequickly") in the Fire Protection career field. One person suggested thatcolor perception might be an important consideration in the Fire Protec-tion specialty, especially for reading hazard codes. On balance, the in-terviewees were largely of the opinion that the 13 abilities included in thetaxonomy were a complete and satisfactory listing of perceptual/ psycho-motor abilities involved in job performance in their career fields.

2A strength requirements survey was taking place at about the same

time as the present work.

93

4.1.

C41EO ICAL

PERCEPTUAL/PSYCHOMOTOR ABIUTY

40.

FD = Finger Dexterity PM = Position MemoryMD = Manual Dexterity AD = Auditory DiscriminationCP = Control Precision AM = Auditory MemoryRC = Rate Control CLP = Clerical PerceptionVM = Visual Memory DP = Depth Perception

VSA = Visual Speed and Accuracy DA = Divided AttentionKM = Kinesthetic Memory

Figure 3-7. Involvement of perceptual /psychomotor abilities inFire Protection career field.

94

40.

wD ft C w V"A M * ML W a^

PERCEPR1AL/pSY040M0IR ANUTY

FD) = Finger Dexterity PM =Position MemoryMD = Manual Dexterity AD = Auditory DiscriminationCP = Control Precision AM =Auditory MemoryRC = Rate Control C LP = Clerical -PerceptionVM = Visual Memory DP = Depth Perception

VSA = Visual Speed and Accuracy DA = Divided AttentionKM = Kinesthetic Memory

Figure 3-8. Involvement of perceptual /psychomotor abilitiesIn Munitions Maintenance career field.

95

Several interviewees alluded to the possibility that differentamounts of an ability could be required at different locations. For ex-ample, Fire Protection personnel may have more occasion to use the"Divided Attention" ability at large air bases because of (a) the morecomplex communications systems at such a location and (b) the greaternumber of radios being attended at the same time. A Munitions Main-tenance interviewee suggested that the amount of "Manual Dexterity"couli vary from base to base, depending on the amount of heavy versuslight munitions that were handled. This suggests the need for includinga variety of bases in surveys such as the present one. The intervieweeswho made these suggestions all have been assigned to other bases priorto their current assignment. Consequently, their suggestions were,more than likely, based on personal experience rather than conjectureor supposition.

An evaluation of the perceptual/psychomotor definitions (ques-tion 6) indicated that all the interviewees understood the definitions ad-equately and were able to relate them to the tasks performed in theircareer field. One interviewee suggested that the examples offered inthe definitions list might be more meaningful if they pertained to motorvehicles, such as cars and trucks, rather than aircraft.

The interviewees were presented with a list of 10 adjectival state -

ments (question 7); five were positive statements and five were negative.The statements appeared in a random order in the list. The intervieweeswere asked to select any two statements which best described their opin-ion of the category scale and another two statements which best describedtheir opinion of the magnitude estimation scale. The distribution of re-sponses, by career field and by supervisor-subordinate, indicated suchminor differences that the responses were grouped (across both special-ties and level) for summary purposes. The data indicated a positive dis-position toward the category scale. A total of 96 responses was record-ed (four responses from each interviewee, two for each type of scale).Of the total number of responses, 54 (56%) were positive, while 42 (44%)were negative. Of the positive responses, 38 (70%) favored the categoryscale over the magnitude estimation scale. Similarly, of the negativeresponses, 32 (70%) were more negative toward the magnitude estimationscale. An alternative view of these data is that when considering onlythe category scale evaluations, 38 (79%) of the responses were favor-able; when the evaluations of the magnitude estimation scale were an-alyzed, 16 (33%) were positive. On the basis of response frequenc ,the three statements that best characterize the opinions of the categoryscale, respectively, were "easy to use," "easy to interpret," and "agood approach. " Conversely, the three most frequently selected optionsdescribing opinions of the magnitude estimation scale were "difficult touse," "provides inadequate information," and "poor approach."

96

The interviewees were asked to evaluate the difficulty they hadin using the two scaling approaches (question 8). They selected one op-tion from five, which best described their opinion on the difficulty is-sue. The responses on the difficulty issue appear to mirror those giv-en in response to the preference question. Approximately twice as manyinterviewees (21 versus 12)rated the category scale as easier ("moder-ately" or "very"). In a similar vein, five times as many interviewees(9 versus 2) rated the magnitude estimation as difficult. The remainingresponses fell in the "neither difficult nor easy" option. However, ingeneral, both scales were considered somewhat easy to use. Threetimes as many of the interviewees (33 versus ll)selected the "very easy"or "moderately easy" options rather than the "difficult" options for bothscales.

Several suggestions were made by the interviewees regardingthe two types of scales. The comments shed additional light on the is-sue of the type of scale which is most acceptable from the point of viewof the respondent:

* It is difficult to use percentages [magnitude es-timation scale ].

0 There are too many choices to consider [magni-tude estimation scale].

0 I can't nail down an answer and be exact with themagnitude scale.

* Don't include the percentages in the categoryscale - it's confusing and too much like the mag-nitude scale.

An open-ended question was asked in order to obtain suggestionsfor improving the final data collection instruments. Some of the sugges-tions were:

* Use only Part I [amount]. I got confused be-tween te two concepts and had to concentratevery hard on the quality of performance section.

* Section II [performance quality variability] wasvery difficult. I had to read it twice to under-stand it.

* It was difficult at first. But after the second col-umn it was easy.

* Place the scales on the pages that have to be filledout.

97

" Make the examples in the definition list specificto my career field.

" Don't use perceptual/psychomotor. Use termslike skill or ability.

The procedures employed in the administration of the formsand the data collection instruments themselves including the definitionsof taxonomic classes) were largely acceptable to the pretest sample ofrespondents and appeared to require only minor modifications to im-prove clarity and meaningfulness (e. g., wording).

Discussion of Pretest Results

Probably, the major purpose of the pretest was to determinewhether or not the task list approach represents a workable techniquefor collecting taxonomic class involvement information for Air Forcespecialties. From the overall point of view, it seems that the techniquewas successful for acquiring the required data. The supervisors andthe subordinates who were involved were able to understand the taxono-my, the scaling procedures, and the methods for completing the forms.

Other issues were also involved in the pretest investigations.The first of issues related to the sensitivity/discriminating power of thescales/taxonomy, There was some indication that the range of responseswas restricted--especially relative to taxonomic classes. This restric-tion may be a true representation of the different ability requirementsor it may represent a scaling problem. The safer course seemed to beto accept the latter explanation and to make some modifications in thescaling procedure for the subsequent large scale data collection effort.On the other hand, when tasks were judged across taxonomic classes,the range restriction was not so pronounced. This finding supports theuse of the task approach that was adopted at the outset.

A second specific issue involved the independence of two consid-erations: the amount of a taxonomic class required for task performanceand the performance quality variability produced by the class. The re-sults indicated a moderate association between the results yielded bythe two factors. There was some indication that the subordinates werebetter able than their supervisors to distinguish between the amount andthe performance quality variability factors. However, the two factorsseemed to be sufficiently independent to warrant continuation of the useof both factors during the extended data collection planned for the nextresearch phase.

98

There was evidence supporting the test-retest reliability of thetechniques and the interrater reliability seemed acceptably high. Thepretest data, in this regard, support the use of the methods/ techniquesof tht- oretest during later study phases.

Some indication was shown of a statistical relationship amongthe 13 taxonomic classes but the associations were, for the most part,not strong. Moreover, the respondents said that they had little difficul-ty in understanding or employing the taxonomy. With this, as well asthe obtained reliability, in mind, revision of the taxonomy did not seemwarranted at this juncture.

The issue of the relationship between estimates made by super-visory as compared to subordinate personnel was investigated from anumber of points-of-view. In general, there was a close relationshipbetween the taxonomic information yielded by supervisors and by sub-ordinates whether type of scaling procedure, career field, taxonomicclass, or rating factor was involved. The overall indication seemedto be that the data yielded by the two personnel levels will be essential-!ly equivalent and that either or both subject groups could form the basisfor the data to be collected in subsequent study phases.

Similarly, there seemed to be little effect of type of scaling pro-cedure on the emergent data regardless of rating factor. However, thecategorical scaling procedure seemed preferred over the magnitude pro-cedure by the repondents. Accordingly, it seemed that the categoricalprocedure should be employed in subsequent data collection efforts.

All of this suggested, that with some modification in wordingsand scaling, judgments possessing adequate quality could be obtainedin subsequent work phases.

99

7'

IV. PERCEPTUAL/ PSYCHOMOTOR REQUIREMENTS

OF 35 AFSCS

With the results of the pretest of methods and procedures onhand, a major effort to acquire information about the perceptual/ psycho-motor requirements of Air Force specialties was instituted.

Changes in Forms and Procedures

On the basis of the pretest results, it was decided to employthe categorical rating procedure during the major effort. This scaleseemed to yield data of equivalent quality, as compared with the magni-tude estimation procedure, and to be easier and more acceptable to therespondents. However, in an effort to increase the range of responses,two additional categories were added to the scale -- extending it from afive category to a seven category scale. Additionally, and for the samepurpose, the percentage values contained in the definitions were deleted.Otherwise, the methods and procedures of the data acquisition remainedthe same as for the pretest.

Sample

Career Fields

A sample of 35 Air Force career fields was selected for inclu-sion in the present work. Tro obtain this sample, the Airman Classifica-tion Manual (AFM 39-1) was initially reviewed to develop an ad hoc listthat would contain career fields in which different types of perceptual/psychomotor requirements seem needed (e. g., manual dexterity, visualmemory, clerical perception). T[his list consisted of 75 career fields.

All specialties which did not possess an Occupational Survey Re-port task listing were eliminated from further consideration. Trhe finalsample was selected from the remaining set of 55 on the basis of apti-tude requirements for entry into the career fields.

The four aptitude areas used by the Air Force to group careerfields are mechanical, administrative, general, and electronics. 'Theaptitude minimums for each career field were listed hierarchically with-in each of thle four aptitude areas. The distribution was then divided in-to three approximately equal segments. '[he highest group of miinimumiaptitude scores was labelled H (high), the middle group wa8 labelled M(mnoderate) and the lowest group was labelled L (low). 'l'he arrangemfienltOf (-utOft Points of aptitude scores, for each aptitude area was:

101

Mechanical

High: 60 or aboveMedium: 50-59Low: 40 or below

Administrative


General )High: 70 or aboveMedium: 50-69Low: 40 or below

Electronics


The final selection of career fields is presented in the matrixshown as Table 4-1. The career fields sampled reflect an equal repre-sentation of each of the aptitude areas and of the high, medium, and lowcutoff points on the aptitude score requirements. Where more than onecareer field was available for selection, that career field was selectedwhich had a larger number of persounel assigned. At the time of thesample selection, the range of personnel (including the skill levels 3through 9) for the career fields listed was 527 to 28, 337. The careerfields in Table 4-1 were identified for the 3-and 5-skill levels.

Air Force Bases

Following the career field sample selection, the next sample tobe selected was Air Force bases. The intent in sampling bases was toconduct surveys at a variety of bases so that geographic and situationaldifferences, if any, could be represented in the ultimate data. A secondconsideration in the air base selection was the air base's major com-mand, e.g., Strategic Air Command (SAC), Tactical Air Command(TAC), or Military Air Command (MAC). Fable 4-2 shows the Air Forcebase sample.

102

Table 4-1

Sample of Air Force Career Fields byAptitude Score Minimumus

HihMedium Log

Mechanical 46110 (Munitions Main- 60SX0 (Air Passenger 603X0 (Vehicle Opera-tenance) and Air Cargo Special- tar/Dispatcher)

ist)462X0 (Weapons Mechan-ic)

5114X0 (Aircraft Load- 63110 (lueis Specialist)master)443X0 (L(G( 25 Missile 552X0 (Carpentry andMechanic) masonry Specialiat)A

Administrative 651X0 (Procurement 29313 (Radio operator) 70210 (Airman Adminis-Specialist) tration)

70110 (Chapel Manage- 732X0 (Personnel)ment)

705X0 (Legal Services) 272X0 (Air Traffic 611X0 (Supply Services) 9Control OperatorTechnician)

27110 (Airport AirOperations)

General 204X0 (Intelligence 645X0 (Inventory Man- 571X0 (Fire Protection)Operations and Imagery agement, MaterielInterpretation) Facilities and Supply

Systems)25110 (Weather lore- 902X0 (Medical Services) 231X1 (Graphics) 9

case)791X0 (Infocmation 981X0 (Dental and Pre- 231X2 (Still Photo-Specialist)C ventive Dentistry graphic)

Technician)D

Electronics 316X00 (Missile Elec- 541X0/G (Missile Fe- 423X0 (Aircraft Elec-tronic Ejuipment Spe- cilities) trical Systems Special-

cialist)L ist)325X1 (Avonic Instru- 542X0 (Electrician) 42313 (Aircraft Fuel 9ment Systems Special- Systms Mechanic)ist)

304X4 (Crourd Radio 54212 (Electrical 426X2 (jet EngineEquipment Repair) Pawer Production) Mechanics)

11 12 12 35

*Missing cell-uo task list available f or additional career fields

Air Force Designations: A. Carpentry Specialist and Masonry SpecialistB. Changed to Weather SpecialistC. Information/Historian and Radio/Television boradcastingD. Dental Specialist and Preventive Dentistry SpecialistE. Missile Electronic Equipment Specialist. G and H Shreds,

Missile Systs Aalyst. G Shred

103

'Fable 4-2

Air Force Base Sample

Air Base Location Function

Davis-Montban Arizona SAC

Nellis Nevada TAC

Little Rock Arkansas TAC

Pope North Carolina TAC

Hurlburt Florida TAC

Bergstrom Texas TAC

K.I. Sawyer Michigan SAC

Grand Forks North Dakota SAC

Malmstrom Montana SAC

McGuire New Jersey MAC

104

Respondent Sample

The respondent sample selection was considered next. On thebasis of the results of the pretest, both supervisor (E-6 and above) andsubordinate (E-5 and below) job incumbents were selected for inclusionin the sample. Selection was from a central locator file. For the incum-bents to be sampled at a base, the base was supplied with a list whichcontained the incumbents' names, grades, and career fields (in a numner-ical code). At the outset, the goal was to select a total of 15 supervisorsand 15 subordinates in each career field. When a selected job incumbentwas not available, the designated project officer at each base was re-quested to make an equivalent substitute. If an equivalent substitutioncould not be made, then a reasonable substitute having the same AFSCas the originally selected incumbent was requested. In some cases,substitution was not possible.

Task Sample

The task list included in the survey form for each career fieldwas designed to be both representative of the tasks performed on thejob and to be of a length which would not overburden the respondents interms of the time required for completion. Completion time informa-tion was taken for each airman who participated in the pretest. On thebasis of these data and a decision that about 2. 5 to 3. 0 hours represent-ed the maximum length for any data collection session, a 60-task listlength was selected. A random selection method was used to develop,for each career field, a list that would be representative of and general-izable to the total AFSC. A table of random numbers was used to com-plete the random selection process.

Tasks were selected from the task listing, provided by the OSRs,for the 5-skill level of each career field. Since the duty categories ofsupervision, planning, and training involve functions often assigned tohigh level, senior incumbents, tasks falling in these categories werenot included in the sampling process. Each of the remaining tasks list-ed in the OSRs was numbered and a table of random numbers was usedto identify those tasks for inclusion in the final list for a career field.Accordingly, each final task list consisted of a set of operational activ-ities common to the AFSC in question.

A downstream problem that could result from the procedure wasthat if changes were introudced in the AFSC (between the time of devel-opment of the OSR and the present work), the final list could then con-sist of tasks which are no longer performed and are perhaps unfamiliarto some responding airmen. To accommodate the respondents who

would, as a result, be unfamiliar with certain tasks, a "Don't Know"' re-sponse was included in the response options.

105

Data Collection Instruments

Content of Data Collection Forms

As stated above, for this data collection effort, several modifi-cations were made to the pretest forms on the basis of the pretest in-dications. Thses modifications involved wording changes made in theinterest of improved clarity, the exclusion of the magnitude estimationscaling approach, the use of a "Don't Know" response category insteadof the "N" (not performed in the squadron) response, and the extensionof the category scale from five units to seven intervals.

The scale of values used in the final survey instruments rangedfrom "1 "(very little), through "7" (very high), with the scale value "4"(moderate) anchoring the midpoint. 3 The scale was presented to the re-spondents in the survey form instructions and also on a card, which t,.erespondents kept before them as they completed their form. Exhibit 4-presents the scale card use(. by the respondents when they complettdthe amount and the performance quality variability sections of the sur -vey form.

The response sheets were identical for both parts of the form.The same format was used as was employed in the pretest. The 60tasks were arranged vertically along the left side of a page, 20 tasksper page. The respondents had a total of 1, 560 judgments to make: 60tasks x 13 abilities x 2 survey form sections.

Demographic Information

The cover page of each data collection form was coded with theappropriate AFSC. The cover page also asked for the name, rank,squadron, and location of the airman completing the form. Additional-ly, the respondents indicated how long they have been in their careerfield and the number of years they have been in the Air Force.

Procedures

Preparatory Step

A letter of introduction prepared by Air Force Human ResourcesLaboratory, Brooks Air Force Base was forwarded to each base severalweeks prior to the survey start. A copy of this letter is presented asAppendix F to this report.

3 The NR response ("ability is not requ.red in duty performance") was

entered as a zero in the subsequent data analyses.

106

0-VFW-

44Aj- 1 -4 4Ae .441 10 0 H)) H 00-

>~U

~' > .-4U W0 '41

-H w ) X) Z) U a) 4 ) - w wkf*- U)4- p $ -4 k p 0

41 r- r- 4 -4 -4 -4 44 CL:3 w a 4) U) ) 4) U) Q 0) [a ) W 0 ) W020) 4.4

'o.4 Ou Q0 4 q)J 0U QU )0 vu0 w U-1 4

z) c00 0 Id 0 0 )0 V )0 0 0 Z010

.0) V' 0)W t U M ac w a 4 A' 0 ) 0)W (1) 4-'.4 a 0)) 0 0 4) )) 0 q ) 4 004)0) 0Q)0) 0CO 1-i 1- 4-4 CL 4444 13. 44 " 0. 44 4A W 444 " 44 4-4 0. ~

:3 - 0 0)r4j a) -4 41 ")~'J A 0)..4 0) '4 4I Q) .41J 0~~

01) u)0 ) ) 00)LW ). 44 w. 444 4-4 44- W 444 W 44 L4- $4 44-i4W

ci rO .. 4 0 0 4 0043.w 00 $-4 00 k 0 0w 0 0 4 0 .0 C

to -4)c > 0) u 0) P") Q)) u0 U > > ci

0 -4."1 c a 14CO tvr -f Cc 0 a9 - 0 Z - coS0 .- 4.0=.-4'44 4.4 '.4 COC0 m Q0 (d Q 0 w u0 00 u0 000u co bo 4 0$.4 V: .44 0a) : 4Q) E : (4) a : :3 a0) : a0) :j -A.0 C)0) :3 CO 1.0 ( a't 0~co 0tr . 0 0oCd V 0 co v o f

0-41

0

z) >%p N

w. 4,q 4

00

04) a)N

'.4~$ -H40)0

414 t0' 'n -.4 4- ' H

p.:% ) W) w . 10 1- 4

41 0) g - 0 ) 0' Z $.A)Cl'.4 0 . a' -,4 :3 (d-4 '-H

9:' 0) a) 0) '.i Q)% 4W- 1 ~ 0 0) u 0-a 0 w0 u- u.

.0) 00% 4.4 4 4-i) 0) e 44.' W. R) 4)0 M '- 4 l 4 0 0

r4 0 00 -4 '.0 . *N 0

4 44 4 44 4- 4 0' >'J COW44 -A40) 4.. ) 4.4 44 w0 10 1)0)k

4) :3 a ) 0 s-I t7 0 04M -t U O t

107

Within two weeks after the introductory letter was transmitted,the appropriate air base was contacted and arrangements were complet-ed for the survey completion. Survey completion at each air base tookfrom two to five days and the same procedures were employed as in thepretest. Data collection sessions were of the group administration na-ture and the sessions were scheduled at the convenience of the air basesinvolved.

Administrator Instructions

The same set of administrator instructions as used in the pre-test were used in administering the final forms. The only change intro-duced was in the sixth point of step 6 (explain the respondcnt's task).This instruction read as follows:

There are yellow cards contained in each bookletthat have the scale values you are to use when mak-ing your judgments (demonstrate). The values rangefrom 1 to 7. If you are of the opinion that a partic-ular skill is not required in the performance of atask, then you would enter "NR" in the appropriatebox (demonstrate). There is also a don't know re-sponse which you can use if you really do not knowanything about the task. Use the "DK" responseonly if you know nothing about the task. Make everyeffort to enter a judgment, even if you don't per-form the task yourself and never did. We are notinterested in what you personally do on the job--only in what you know about the performance of thetasks on the list. Your knowledge of thses taskscould have come from your training, observation,or past performance. Try to respond in every in-stance with a scale value.

Demographic Description of Sample

Survey forms were completed by 808 airmen in 35 Air ForceSpecialty Codes (AFSCs) in four aptitude areas (mechanical, administra-tive, general, and electronics), sampled at 10 Air Force bases. Thefinal sample represented 80% of the 1008 airmen identified for partici-pation in the survey.

The percentage of completed survey forms from the 10 air bas-es ranged from 64 to 89 with a median of 77.50. With the exception ofan "oversampling" in Carpentry and Masonry Specialist and the

108

coincidence of actual size with requested sample size in ProcurementSpecialist, "undersampling" in the remaining 33 AFSCs ranged from 7to 50 percent with a median of 22 percent.

The eight AFSCs in the mechanical aptitude area contributed 181out of 808 (22%) airmen to the total sample; the nine AFSCs in the ad-ministrative aptitude area contributed 206 (25%) airmen to the total sam-ple, the nine AFSCs in the general aptitude area contributed 203 (25%)airmen to the final sample; and nine AFSCs in the electronics area (con-tributed 218 (27%) airmen to the final sample. Thus, the four aptitudeareas were approximately equally represented in the total sample. Eachof the four areas contributed about one quarter of the total final sample.

Supervisor Description

Three hundred and seventy supervisors (46%) participated in thefinal survey. The frequency of supervisors in the 35 AFSCs in the fouraptitude areas together with median and modal coded number of years inan AFSC and years in the Air Force are reported in Table 4-3. The fre-quency of supervisors in the final sample ranged from 1 to 15 in an AFSCwith a median of 11. 14.

Number of years in an AFSC for supervisors was coded as fol-lows: . Less than 1 year to 1 year.

2. More than I to 4 years.

3. More than 4 to 7 years.

4. More than 7 to 10 years.

5. More than 10 years.

The median of coded number of years (calculated from class intervalsof scale values and converting to years, as indicated above) in an AFSCranged from 2. 25 (more than I to 4 years) to 5. 00 (more than 10 years).In 25 (71%) of the AFSCs, the median number of years in an AFSC wasmore than 10 years. The modes of coded number of years in an AFSCranged from 2. 00 (more than 1 to 4 years) to 5. 00 (more than 10 years).In 31 (89%) AFSCs, the mode of coded number of years in an AFSC wasmore than 10 years.

Number of years in the Air Force for supervisors was coded ina manner identical with the code for years in an AFSC. The median ofcoded number of years in the Air Force ranged from 4. 00 (more than7 to 10 years) to 5. 00 (more than 10 years). In 33 (94%) AFSCs, themedian of coded number of years in the Air Force was more than 10

109

ATable 4-3

Frequency of Supervisors in the Final Sample in 35 AFSCs inFour Aptitude Areas with Median and Mode of Years in their

Specialty (AFSC) and Years in the Air Force (AF) +

Years in Years inAFSC AF

AFSC n Mdn Mo 1dr. Mo

Mechanical

Munitions Maintenance 12 4.50 5.00 5.00 5.00Weapons Mechanic 12 4.64 5.00 5.00 5.00Air Passenger and Air Cargo Specialist 2 4.50 4.00;5.00 4.50 4.00;5.00Aircraft Loadmaster 11 4.95 5.00 5.00 5.00LGM 25 Missile Mechanic 13 2.75 2.00 5.00 5.00Vehicle Operator/Dispatcher 11 4.58 5.00 4.81 5.00Fuels Specialist 13 4.96 5.00 4.96 5.00Carpentry and Masonry Specialist 1 4.00 4.00 4.00 4.00

Administrative

Procurement Specialist 14 4.50 5.00 5.00 5.00Chapel Management 11 4.71 5.00 4.95 5.00Legal Services 12 4.00* 4.00*;5.00* 5.00* 5.00*Radio Operator 2 5.00 5.00 5.00 5.00Personnel 12 4.50 5.00 5.00 5.00Air Traffic Control Operator Technician 10 4.94 5.00 5.00 5.00Airman Administration 15 4.96 5.00 4.96 5.00Supply Services 10 4.67 5.00 5.00 5.00Airport Air Operations 10 4.60* 5.00* 4.99 5.00

General

Intelligence Operations and Imagery Interpretation 8 4.93 5.00 4.03 5.00Weather Forecaster 10 4.67 5.00 5.0 5.00Information Specialist 9 3.25 5.00 5.00 5.00Inventory Management, Materiels Facilities andSupply Systems 13 5.00 5.00 5.00 5.00

Medical Services 10 4.94 5.00 5.00 5.00Dental and Preventive Dentistry Technician 11 4.89 5.00 4.95 5.00Fire Protection 11 4.95 5.00 4.95 5.00Graphics 8 4.83 5.00 5.00 5.00Still Photographic 11 4.89 5.00 5.00 5.00

Electronics

'fissile Electronic Equipment Specialist 12 2..2* 2.00* 5.00 5.00Missile Facilities 12 2.25 2.00 4.95 5.00Aircraft Electrical Systems Specialist 14 4.92 5.00 4.)2 5.00Avionic Instrument Systems Specialist 12 4.95* 5.00* 5.00 5.00Ground Radio Equipment Repair 9 4.93 5.00 5.00 5.00Electrician ii 4.83 5.00 *.95 5.00Electrical Power Production 8 ..83 5.00 5.00 5.30Aircraft Fuel Systems Mechanic 15 4.82 5.00 4.9b 5.00Jet Engine Mechanic 15 5.00 5.00 5.00 5.00

.n 370

'One respondent did not check a category on the background cover sheet.

Years in Air Force are necessarily more than vears in specialty for a number of reasons.

110

years. Similarly the modes of coded number of years in the Air Forceranged from 4. 00 to 5. 00. In 34 (97%) AFSCs, the mode of coded num-ber of years in the Air Force was more than 10 years.

Subordinate Description

In the final survey, 438 subordinates (54% of the total number ofparticipants) participated. The frequencyof subordinates in the 35AFSCsin the four aptitude areas together with the median and mode of codednumber of years in an AFSC and the years in the Air Force is reportedin Table 4-4. The frequency of subordinates in the final sample rangedfrom 5 to 20 in an AFSC with a median of 12. 08.

Number of years in an AFSC for subordinates was coded in amanner identical with the code for years in an AFSC for supervisors.The median of coded number of years in an AFSC ranged from 1. 88(more than I to 4 years) to 4.50 (more than 7 to 10 years). In 11 (31%)AFSCs, the median of coded number of years in an AFSC was more than1 to 4 years. Similarly, the modes of coded number of years in an AFSCranged from 2. 00 (more than 1 to 4 years) to 5. 00 (more than 10 years).In 21 (60%) AFSCs, the mode of coded number of years in an AFSC wasmore than I to 4 years.

Number of years in the Air Force for subordinates was coded ina manner identical with the code for years in an AFSC. The median ofcoded number of years in the Air Force ranged from 2. 38 (more than Ito 4 years) to 4. 50 (more than 7 to 10 years). In 23 (67%) AFSCs, themedian of coded number of years in the Air Force was more than 4 to7 years. Similarly, the modes of coded number of years in the AirForce ranged from 2. 00 (more than I to 4 years) to 5. 00 (more than 10years). In 13 (37%) AFSCs, the mode of coded number of years in theAir Force was more than 1 to 4 years and in 14 (40%) AFSCs the modeof coded number of years in the Air Force was more than 4 to 7 years.

Accordingly, the number of supervisors and subordinates in thefinal sample was approximately equal. But, as a group, the supervi-sors possessed more years than the subordinates in an AFSC and in theAir Force, as could be expected. Moreover, as a group, the supervi-sors were more homogeneous than the subordinates with respect to yearsin an AFSC and in the Air Force.

Except for Carpentry and Masonry Specialist, Air Passenger andAir Cargo Specialist, and Radio Operator, in 24 (69%) AFSCs the discre-pancy between sample size of supervisors and subordinates did not ex-ceed three airmen.

111

.J

Table 4-4

Frequency of Subordinates in the Final Sample in 35 AFSCs inFour Aptitude Areas with Median and Mode of Years in their

Specialty (AFSC) and Years in the Air Force (AF) +

Years in Years inAFSC AF

AFSC n Mdn Mo Mdn :o

Mechanical

Munitions Maintenance 14 2.67* 2.00* 2.67* 2.00*Weapons Mechanic 15 2.80 2.00 2.83 2.00Air Passenger and Air Cargo Specialist 11 2.38 2.00 2.38 2.30Aircraft Loadmaster 11 2.63 2.00;3.00 3.25 3.00LGM 25 Missile Mechanic 10 2.67 3.00 3.40* 3.00*Vehicle Operator/Dispatcher 12 4.50 5.00 4.30 5.00Fuels Specialist 13 3.71 4.00 3.71 4.00Carpentry and Masonry Specialist 20 2.71" 2.00*;3.00* 2.83 3.00

Administrative

Procurement Specialist 16 2.50 2.00;3.00 3.00 3.00Chapel Management 11 2.67 2.00 3.00 2.00Legal Services 1L 1.93 2.00 4.17 5.30Radio Operator 13 3.58 4.00 3.58 4.30Personnel 13 3.25 2.00 3.b7 5.00Air Traffic Control Operator Technician 17 2.15 2.00 2.75 2.200Airman Administration 10 3.00 2.00 3.25 3.00Supply Services 11 3.33 3.00;5.00 4.00 5.00Airport Air Operations 5 1.88 2.00 3.75 4.00

General

Intelligence Operations and Imagery Interpretation 11 2.68 3.00 3.57 3.00;5.00Weather Forecaster 11 2.81 3.00 2.92 2.00;3.00;4.00Information Specialist 12 2.57 3.00 3.00 3.00Inventory Managenent, Materiel Facilities andSupply Systems 15 3.40 4.00 3.75 4.00

Medical Services 11 2.92 3.00 3.00 3.00Dental and Preventive Dentistry Technician 15 2.86 3.00 2.90* 2.00*;3.00*Fire Protection 12 3.50 2.00;4.00 3.50 2.00;4.00Graphics 11 2.40 2.00 2.80 3.00Still Photographic 14 3.00 3.00 3.50 5.00

Elect ronics

Missile Electronic Equipment Specialist 11 2.S6 2.00 2.57 -00Mismile Facilities 12 2.21 2.00 3.00 3.00Aircraft Electrical Systems Specialist 14 2.50 2.00 2.b7 2.00;3.C10Avonic Instrument Systems Specialist 11 3.08 3.00 3.)8 -.00Ground Radio Equipment Repair 12 2.50 2.30 -'.83 -. 0Electrician 16 2.50 2.00 3.83 5.00Electricial Power Production 13 3.00 2.30 3.00* -.20*Aircraft Fuel Systems Mechanic 12 2.50 2.00 2.30 2.00Jet Engine Mechanic 9 3.00 3.30 3.13 3.00

::n -.38

*One respondent did nct check a category on the backgrcund --ver sheet.

-Years in Air F-rce are necessarilv more than -ears in specialtv fc! a r.urber reasuns.

112

Results

In the 35 AFSCs, 370 supervisors and 438 subordinates judged the

amount of 13 perceptual/psychomotor ability classes relative to the per-formance of 60 tasks in their AFSC, on a seven-point rating scale. Al-

together. 780 judgments were made by each airman in this aspect of the

survey.

Taxonomic Class Means--Amount

The 60 judgments by a supervisor were summated for an ability

and reduced to an arithmetic mean for each supervisor in the 35 AFSCs

to generate an individual mean for each supervisor. Thirteen sets of

individual means resulted for each supervisor in each of the 35 AFSCs.

(The total number of individual means calculated was 4, 810). These

were further reduced to a group mean for the sampled supervisors in

each AFSC. Thirteen sets of group means resulted for the supervisorsin each of the 35 AFSCs. The total number of group means for super-

visors calculated was (13 x 35) 455.

In a like manner individual means and group means of the amount

of a perceptual/psychomotor ability by sampled subordinates were de-termined. There were 5,694 individual means and 455 group means cal-

culated for subordinates.

Finally, the 60 judgments of amount of a perceptual/psychomo-

tor ability by all supervisors and all subordinates in an AFSC were sum-mated for an ability and reduced to a grand mean to generate a combined

mean for supervisors and subordinates combined. The total number of

such means calculated was 455. These means, combined for supervi-

sors and subordinates are reported in Tables 4-5 through 4-8, accord-

ing to the four aptitude areas.

The mean ratings reported in Tables 4-5 through 4-8 generally

cluster about the low end of the seven-point rating scale. In 368 out of455 (81%) of the instances, the combined mean is below 3.0, -- "some"

amount of the designated ability was judged to be involved in performingthe tasks in an AFSC. This trend is consistent with the pretest findings.

In 416 out of 455 (91%) comparisons of the group means for sam-

pled supervisors and the group means for sampled subordinates with re-

spect to judgments of amount of the 13 perceptual/psychomotor abilities,

differences of less than one point on the rating scale were found. Twenty

out of 455 (4%) differences larger than one point on the rating scale clus-

tered around visual memory, visual speed and accuracy, and position

113

.8 o IT 'Goo 100% 1118 wg@ Cc" c 0'- @"' 1

*0 ~~~L l -~r~~..o -o

f4 - c Ne 4 r4d t4 nf4 ld -

N O4 - -w 0 C0 f4 -o LIN '

rA U

-- --, -- ell 00 L- 'T2UCU 0 )

0.e: o . . .t

0s e4 V%. a, e4N ODW 0. ede C% r4 -am-0

c4 -- a

49N 94 O o n- d '' .. .

1-1 C c4. r4e mi 49 " 4 f984 r4ed edd 4 0

44 CO C0~On .. ~.~

lo ccO .'4 lo q f40 m.. .N N me 0

0 v4ded m9- "9N N N4 m 4 v%- edd 4

bO- .. O fle - -- -- e n- a: n eb. ul4

-4 C,.0 0! 0.( N0 Wed N0' GD el C rN N . e

VS. r4c - --C 4e

z0 OD cN m05 m0 ... ( -. o .40 OS .4 0

0 ~ ~ ~ ~ c Ic- (9 NO W. 0'- fi. 50 5 ONeIn t4N r4S N&4e 'S .- Nd (

ba S.4

0

'0 N (9 N (9 9 '0 - - c

ul )

n .4 44

I4- Ua4 cm a L-

114

4) 0

C0- 00 0- r4' 0 - - - 0- 0-

0 > 10 cc: 1oO .r' . D wo 'r.~' o

o cd 0- 0;- 00 -~0- N 0- - 0- 0

C.) ": cc M% -1 l0 CN 'co WC go~ O0 ~ 'mu Z

10t z~ 0 CO' OD .. 0- a.' '0 0 -0~ O

4) 7 CC

$4.

CN4f U" . 2N O 0 0 ' a, = co" e c

%09C GoN 01N NO CN Do -N CN

orz U PU o00C 0

SON CN CC I c N 11) 0-14'N. N0

0 0U,. 4D 0 0.l 0-0 In' -- .- C 00 --

0.~O CyN N .. o-CdC

0o >

COCEnS N C' N ~ ' - N C C C

o NC N CS . 5 *N w CN - N N

0 ~f &I ~ ZI0 aIl2 t~j 0. ~0> Q

U)4) - .

r N >0 CC .- N ~ - " 0 C

- cN N7 N N N

- N C

VtcU 0j r CL

'44 a i-4 - - C

CU~ a CU15

W ~ , ' co 7,- m 0.' 4* 0~ w n

00 0 0- 00 - - - - - -

0 - p -4 0. CNN -0 cc *

0. ~ ~ 1 -- 0 0 NN

O caC) 0% T.C InO .W2 go cc- 40 ~ ~

C0' 14 - 1 4 Iz 4 4 4 M.O ~ e ~ C cn

4

GCd

CO 10 vMO ~ ~ N '0 M '0 O

LM 0C osO Z:D-C/2

.0 ccn ,C4 c 0C

r. MCN MO. C4 ON N' Nf OM -e N

00

Ca -7 C7 ca 11 c

0, ,., ., .Z

0-

0'0 00

5 0 It 5

UC -

(U r- 13 -0 w ri0 ,

uJ 1.. r 7 CCL) a tt 0 0 - - 0

>C x; C -

.0

U L .116

0 In M CID ~ C C C. O MC 0

-- NN ~ N-' -4 0- .. - -

1400O C C- MM -0 2. z 00% CO NCc

C! 9 ~ 04 NC N.0 0N 0:

NN N- --. -

00 010 a 1 CO m1.~1. 4i. 0N M C~ 0. .- 00 N M' W

coo ca~0 MO LO 04, 00 cNc N 40 -

U2

401 v)

~~~~1 00 I. C . ' M

0N .. N 0- 0-" e: .-

000

W C

I.,' fO M, r N ccN cc 1' MN7 1M N '

In V41

CM eM MM cc NID cC G.o ID r ... c-.. ~~~4 14 9 ~ 44 e-' -. 0 oN o: C' O sO O -

>4. n

::s ~- MILnl ~ l' c IN X: Lon NI Xo I NN MN Ml NI

'4- W c

M5Si INN NC ON 0 N4 .00 C -N 4- 0)

c , . O O M . - N I .N - SM NM0M N M N M

w tov Le)

oS ~ ~ ~ ' 6 IF r. CL. 6MMS M . O .- 4 ' N - 0v u C,.F~ -- NIN 4N.. M NI NN N -INN NN 1

.. 4 9 rd r

117

W.

memory across the 35 AFSCs. The use of combined means for super-visors and subordinates, instead of individual group means was indicat-ed by the pretest results and further supported by the present data.

In 51 out of 52 (98%) comparisons, aptitude area means for eachability separately differed from one another by less than one point onthe rating scale. For 48 out of 52 (92%) of the aptitude area means, theobtained values were less than 2. 00.

Taxonomic Class Standard Deviations--Amount

A standard deviation (SD) was also determined as a summary in-dex of variability/scatter of the judgments of the amount of each percep-tual/ psychomotor ability over the 60 tasks in the supervisor and subor-dinate subsamples separately in each AFSC. That is to say, the 60 judg-ments by all supervisors in an AFSC were summated for an ability andreduced to an SD to generate a group SD for the sampled supervisors inan AFSC. Thirteen sets of group SDs resulted for the supervisors ineach of the 35 AFSCs. An identical variability determination was madefor the sampled subordinates in each AFSC. Thirteen sets of group SDsresulted for the subordinates in each of the 35 AFSCs. For each sub-sample, therefore, 455 group SDs were calculated.

Finally, the 60 judgments of amount of a perceptual/ psychomo-tor ability by all supervisors and all subordinates in an AFSC were sum-mated for each ability and reduced to generate a SD combined for super-visors and subordinates. The total number of such SDs calculated was455. These SDs combined for sampled supervisors and sampled subor-dinates are included in Tables 4-5 through 4-8 according to the four ap-titude areas.

As in the pretest, the subordinates in the 35 AFSCs were char-acteristically more variable than supervisors in their judgments ofamount of 13 perceptual/psychomotor abilities in the performance of as-signed tasks in an AFSC. Thus, in 12 out of 455 (3%) comparisons ofthe group supervisor and group subordinate SDs, subordinates exceed-ed supervisors by less than one point on the rating scale, in 183 out of455 (40%) comparisons, between one and two points on the rating scale,and, in 260 (57%) comparisons, by more than two points on the ratingscale. It is possible that this finding may be due to the heterogeneityof the backgrounds of the sample subordinates in terms of years in anAFSC and in the Air Force, as compared to the relative uniformity ofthe backgrounds of the supervisors on these dimensions. The subordi-nates, as a group, it could be surmised, may lack a common frame ofreference or a firm base for veridical judgments. Ability and motiva-tional variables, too, may account for the observed differences.

118

Within the administrative, general, and electronics aptitude ar-eas, all aptitude area SDs differed from one another by less than onepoint on the rating scale; and, in the mechanical aptitude area, 3 out of13 (23%) aptitude area SDs differed by more than one-point on the ratingscale. Within the same three aptitude areas, 23 out of 39 (59%) of theaptitude area SDs were larger than 2. 00, and in the fourth, 10 out of 13(77%) of the aptitude area SDs were so.

"High" and "Low" Ability Requirements

The combined means of judgments of the amount of the 13 indi-vidual perceptual/ psychomotor abilities by all sampled supervisors andsubordinates within an AFSC were compared with the AFSC mean acrossall abilities. If an individual mean deviated from the AFSC mean acrossabilities by at least one SD unit above the AFSC mean, it was inferredthat the sampled airmen judged the tasks in an AFSC to require a rela-tively "high" amount of the individual perceptual/psychomotor abilitieson the performance of its tasks. If an ability mean deviated from theacross-ability mean by at least one SD unit below the AFSC mean, itwas inferred that the sampled airmen judged the tasks in an AFSC to re-quire a relatively "low" amount of the individual ability on the perform-ance of its tasks. These relatively "high" and relatively "low" judgmentsof amount of 13 perceptual/psychomotor abilities are reported for 29AFSCs in Table 4-9 according to the four aptitude areas. For sixAFSCs no ability met the plus or minus one SD criterion: Mechanical--M4unitions Maintenance, Air Passenger and Air Cargo Specialist, Vehi-cle Operator/Dispatcher; Administrative- -Radio Operator, and Gener-al- - Fire Protection.

Distinctive patterns of perceptual/psychomotor abilities emergedas associated with the performance of assigned tasks in most AFSCs.Because of overlap of relatively "high" amounts of judged specific abil-ities within and between the four aptitude areas (like visual memory andfinger dexterity in each of the four aptitude areas and clerical perceptionin the administrative and general aptitude areas and position memoryboth in the mechanical and electronics aptitude areas) the pattern of per-ceptual/ psychomotor abilities associated with the performance of assign-ed tasks assumes some importance. Moreover, the patterns listed inTable 4-9 seem to be in common sense accord with actual job re-quirements. For example, visual memory and divided attention wouldbe expected in an Air Traffic Control Operator; finger dexterity andmanual dexterity in a Dental and Preventive Dentistry Technician; fin-ger dexterity, manual dexterity, control precision, visual memory,visual speed and accuracy, and position memory in a Ground RadioEquipment Repair Specialist; and, finger dexterity, visual memory,

119

0 ~00 0 0 CL 0 0 0

0 03

, C CL. I-

o 0 o 0 0 00. 00

r4. .4 .44-

.0. F CL

o4 0 00. 0.0 0

0 u,

o~~~0 00 000oa 00 00a0 a u0

[- CL ~ C, 0 0 C 0w

Cl) :3 C:- . W 4 wd) 0 0 a1w ww 4:w 9:c C Qa c- . ' c

0 4

b.0~0

-Ia~~1 IV 4 .- 4

0~~ CL4'4 4 419 0 t m'- 0 o u 0 c04 cm44 OW mI 0

m .4 a 0 . ,4 . ~ -

-4 ~ ~ ~ ~ X CL> O .C C) >

wol-.CL .- m a 0c~

aO.0 0 1a 0 a 0 a a a .

0 -A* 00C 0 0 .4 000

V0 a4 0 a~ a0 * V

4 .4

0 0.4 0 0 v) 01 05*i 000 0 0O

0W *.* X4 V4 00 4 4

w4 4w 0* o.v L0U0 *0 u > a

w 4 w In0 4 . 0 0

0 i0a 0 0 C u U> 0 0i ..0 w-~5~. S~

00 c) 0 0* 00 0 ou 0 0 0 00a 0 0

0~~~ w 4.4 0) w *4 0 0> 0 . '

0

0. 0 C,4*4*

V0 *4 0 *4 w4 0) *40 4 * 4In) v. w0I .* 0 ua0 0 a. )) w

C4 611 Or * .0 *0 o4w L$ m4 a)-u *4 aL F11. 1 w. )) m

01 0 0 'a 05 m0* m*0 M m 0

0 0 10 m0 v1 r* - *59

C* r X0 0, :3 1)) * . 0*0 0

c *. 0 .01 0 .40 0 00m .0 u0 00 W *0 -4 EV 0. u4 o 0 *0 )Fw a Ww E a 0 1w0 *4 0*. 0 00 C, "~ w . 00

9 ~00 *0 Y 4. F. 5*0* X 01 01 w.w4 00 .4 u*0.4 3.4*4 . a .4* (a . 0 C400 -4 *4g 4 0 - M4* >.

G)~~~~~ v0 14 010 *4 *0 IV4 * 0 * 4014) W ~ WI4 w. *44 a *d *4 n. w0 *4 4 4, *4) 5*

4/1 *0 o 00 08 0I o 0 w~ *i 00n> 0- 0.4 0 S16 00 - r0 Z 4 -V* z0 u4 >4 x0. A

C- U4** 4 * 4 . 4 * -0 *0 0 4. *44H ,4 *4 )W * 0 4* 0 0 * 0o1** ** I 00 *1

00 >.

.4* 'd 00 1 L

-4 *0 c. *4 u *~

*0 a > . 0i >4 *0 *CL q CL *0 *o w

00 *0 n .c4 u 0 *00 o

*4*4.4 f .4 0 -4*0 * a.

-. 4 r0a01

01

u. *0C0 0 .

C: t,* w* 14o . .4 *4 w4 -4 * ) .4 * *4 *40

121

visual speed and accuracy, and position memory in a Weapons Mechanic.Other examples, like Weather Forecaster, Personnel Specialist, Jet

Engine Mechanic, and Aircraft Loadmaster patterns of perceptual/psy-

chomotor abilities support a correspondence between the judgments of

these abilities and actual requirements "on the job" scene. According-ly, a pattern of perceptual/psychomotor abilities, taken with other in-dices of cognitive and personality factors, could be decisive for person-nel assignments.

With respect to the relatively "low" amounts of judged specificperceptual/psychomotor abilities among the four aptitude areas, audi-tory discrimination and auditory memory were consistently judged as

having "little" or "very little" involvement in task performance withinand between AFSCs in the four aptitude areas. Kinesthetic memory,

too, was generally minimized in the administrative and general aptitude

areas--in accord with logical expectations. In a similar fashion, depthperception cannot be expected in Legal Services and in Inventory Manage-

ment, Materiel Facilities, and Supply Systems. As for the patterns of

relatively "high" amounts of judged specific perceptual/psychomotorabilities, patterns of relatively "low" amounts of judged abilities can

be useful in personnel classification. The "low" and "high" patterns,integrated into a unit for personnel classification, should improve the

probability of reducing the incidence of assigning to a specialty personswith abilities not required by that specialty.

Across all AFSCs, Table 4-9 indicates the four abilities mostrequired are: visual memory, visual speed and accuracy, finger dex-

terity, and manual dexterity. The four least required are: auditorymemory, auditory discrimination, kinesthetic memory, and depth per-ception.

Taxonomic Class Means- -Performance Quality Variability

Each of the airmen judged the performance quality variabilityas a function of the 13 perceptual/psychomotor abilities separately for

each of 60 tasks proper for duty performance in their assigned AFSC.

Altogether, 780 judgments were made by each airman during this

phase of the survey.

Individual means for supervisors and for subordinates, group

means for supervisors and subordinates, as well as combined means

for supervisors and subordinates were calculated for the (60 judgmentsof performance quality variability as a function of each of the 13 per-

ceptual/psychomotor abilities in the 35 AFSCs. As previously, there

122

were 5, 694 individual means, 455 group means, and 455 combined means.The combined means for all sampled supervisors and all sampled sub-ordinates in an AFSC are presented in Tables 4-10 through 4-13 accord-ing to the four aptitude areas.

Differences of less than one point on the rating scale were foundin 438 (96%) comparisons of the group means for supervisors and thegroup means for subordinates with respect to judgments of performancequality variability as a function of the 13 perceptual/ psychomotor abili-ties. Eleven (2%) differences larger than one point on the rating scaleclustered around visual memory, visual speed and accuracy, and posi-tion memory across the 35 AFSCs. The use of combined means for su-pervisors and subordinates, instead of a mean for supervisors and amean for subordinates in Tables 4-10 through 4-13, therefore, was againbelieved to be supported.

Group SDs and combined SDs for all sampled supervisors weresimilarly calculated. The combined SDs for all sampled supervisorsand all sampled subordinates in an AFSC are reported in Tables 4-10through 4-13 according to the four aptitude areas.

As previously, the subordinates in the 35 AFSCs were charac-teristically more variable than were the supervisors in their judgmentsof performance quality variability as a function of the 13 perceptual/psy-chomotor abilities in the performance of assigned tasks in an AFSC. In $22 out of 455 (5%) comparisons of the group supervisor and group subor-dinate SDs, subordinates exceeded supervisors by less than one pointon the rating scale; in 299 out of 455 (66%) comparisons, between oneand two points on the rating scale; and in 134 out of 455 (29%) compari-sons, by more than two points on the rating scale.

Aptitude area means and SDs were also computed for judgmentsof performance quality variability as a function of the 13 perceptual/ psy-chomotor abilities in the performance of assigned tasks in an AFSC, asdescribed for the determination of aptitude area means and SDs for judg-ments of amount of each ability in the preceding section. The aptitudearea means and SDs are also reported in Tables 4-10 through 4-13.

In all 52 (100%) comparisons, aptitude area means for each abil-ity separately differed from one another by less than one point on therating scale. Thirty-three out of 52 (63%) of the aptitude area meanswere Less than 2.00.

It was also noted that all aptitude area performance quality var-iability SDs differed from one another by less than one point on the rat-ing scale. In nine out of 52 (17%) comparisons, the aptitude area SD waslarger than 2. 00.

123

y-q

01

4m 0 -- N

) .3 -

--0- 'D- -.

0 000 0-

r. P000 ... --- J N C ~ CJ - N

.4 0

C. - 4

t~ - 0' -30 '- 0 .o l ,

o o-enA

-/ 0r. '3 -0 N '' r.' N' 3 '

-0 8 3 -' ". - .'.0 C *d1

LiA C C. -

0 ~ N N- N N ((N --124

-- GNP

0A0C 0- 0- a - 0a - 00 00 0-

0N 0Y 1N 00CN

4-4

ad 0- C0 -- 0- -N 0- - 000-

.. n

0 -i-

U M 02' '0- O l "' ' 4 ~ 0 O' .. '

CIS 0C 0N0 - C.

0cc w~ 00 00 0-.(

-4.0 C

bi E . . C-

U)U

U C-

U.) Q z C

0- 0-. --- -

a C

CV) Cd CO N~~ ( N C O O

Ct) ZZ

Ci

Oo N U0" I( ( P

CL 0=5UC c >~ 0

V)'.4 <C

aC 0 , - , .I. - C- - C ~ C

125

00 0- 00 00 -- - 0-~ -- 0

I-tO O ' 2.~ :01. -050 2007 z

-- V-4-C - - - -N

r 0,4 0- sN 4'm 00 Go' CIO - C O

>44 .- 55 U.O 1'- 4'- .2 9 4

"-4 14~

00

00 00 0- 00 - 0

4.4 -4V;C8L CU s0 1..t -N 40 05..7 N 0.4 CO5 IT

C14 .0 M .

S: () 01 0- 0 T T:

0 =%

.M~~I z)or. ~ o s- o~ 4.7 w '- os- w~.~ .~*

45 C.N .. 0 . . ~ - C4 0 NO - 0..

0 0

0 ~C00

q. Ec 44 0~ N4. '4 C S 4-i?0 -O SN 4S

ca .d CA

cq~~4 NN sS N N-. N N N NN CN CN 1r4 > m

cd Co .bD-~~~~~~~~()~~~ C40 N 45 CN .S5 .S 5 50 4 ' -0.0 v O 4 .T 445. Q .. O N W.4. 91 7

01U v 0-L,. -N c.. -

.0 U) ~~~~ 5.~~~.40 -- 445.' -54 C N CNi. .. 545 0.CU C -. ~ ~ . o. '-. '. s.= o 0 -e c.o 0 -

0~ cc a Is 0 -- . N N . N --

Q C:5cN 40' c0 aO N 0 ~5 N 54

U) ) ) '50 500 .70 045 -s.. .0 O126-0

NSUU) Z -law

Q) VN 0 0 0 -a -h r

> 0

0 -d. 0 --

0c CI I- cc r -c

Z~~~ C4 -a. 4-e C4~e a'o C..3 4a' N~ NO

w vi- eq CC

aa

0u 0 - - o- - -- -

0cS- L1% C m' n o 0

bjG CU)5 e., O~ 3.s s 00 .45 s * ?'

-Z e4 - - -N - - -~ N.- -- -

C L 7 ,.. c - 1.

41 .A e4~ 544~~ - - .N 'SN .-. N - N N N N C

10.7 L^ e

4-4-

0 ..

t 4 flt ao'4Er tA i'i I(r 1CI

00 '

Cu rn' 0' .- O O 4 f' .' 5 ('

U2U

.4- a 54 w. . . C

00 u0 v z00 X

I- C

Cu~ Cu-CC, It. u. 2- v

6) V - C) r-Z41 -LW~w LI 4.

~ 4)45 U127

54 CU *- Wa -- ap ~ , 4

Within the electronics aptitude area, three out of 13 (23%) aptitudearea means differed from one another by more than one point on the rat-ing scale; in each of the general and mechanical aptitude areas, foul- (31%)differed by more than one point on the rating scale; and, in the adminis-trative aptitude area, five out of 13 (38%) differed by more than one pointon the rating scale. For each of the four aptitude areas overall, all ap-titude area SDs differed from one another by less than one point on therating scale.

"High" and "Low" Performance Variability Classes

The combined means of judgments of performance quality varia-bility of each of the 13 perceptual/psychomotor abilities by all sampledsupervisors and subordinates within an AFSC were compared with theoverall AFSC mean. If an individual mean deviated from the AFSC meanby at least one SD unit above the AFSC mean, it was inferred that thesampled airmen judged the performance quality variability as a functionof the 13 perceptual/psychomotor abilities to be relatively "high" on theability. If an individual mean deviated from the AFSC overall mean by atleast one SD unit below the AFSC mean, it was inferred that the sampledairmen judged the performance quality variability as a function of the 13perceptual/psychomotor tasks to be relatively "low" in the performanceof assigned tasks within an AFSC. The relatively "high" and relatively"low" judgments of performance quality variability as a function of the 13perceptual/psychomotor abilities are presented for 18 AFSCs in Table 4-14 by aptitude area. For the remaining AFSCs, the plus or minus oneSD criterion was not met by any individual taxonomic class.

With the exception of Radio Operator in Table 4-14, the AFSCspreviously reported in Table 4-9 are represented again in 'Fable 4-14.However, Table 4-14 includes 11 fewer AFSCs than does Table 4-9.This suggests (a) some consistency in results and (b) somewhat less dis-criminatory power for the performance quality variability question. Wherean AFSC is common to both tables, there is a "high" correspondence be-tween perceptual/psychomotor abilities previously reported in Table 4-9and again reported in Table 4-14, i.e., if an ability was indicated in Ta-ble 4-9, it is highly probable that it will be indicated again in Table 4-14.This correspondence suggests a perceptiveness of the critical abilitiesrequired in the performance of assigned tasks within an Al,'SC. What-ever is essential for successful execution of a task can be expected todominate observational awareness/alertness. Individual differences inefficient performance, as a rule, are readily communicated and quicklyregistered. Thus, if a "high" amount of a perceptual/psychomotor abil-ity is required for the performance of assigned tasks within an A,'.S(', itwould follow that not all personnel so engaged would be equally prolicji ill

128

o 0 0~ 0

. u u~Cd0 .' 4'

4)40

.4 05 C~

o 4) A44 44

.4 5Q

00 0 w, 000 0

0d 00 04 0 ±~.2~ .0

0. .4 *

00C64U 4

.44 t4 *0

u x4. ~0 444

~' 0 0 -4 '0 00 ~U '

0- S >4 0 56, -. :,

ow o

41 4-0 ..-4)

0.

00.4 0

0S 0M.

-44

040

to U

00

129

0 0 0 0 00 0

0U 00 00 CL

0 'A 0 , 'A

0000 0 060 0 a00 0C-4 4. .0~ -4 .

. .46

0 ~ ~ ~ 00 000 0

04, 04 0404

00 0 000 00 00.

a0'0 14 '0 0

'6 . 6

0

0000 a. aj 0 a.

00 t. 0. 44 W606In 93 a 4 a - a 4 0Q

ca 4)6 64 04 w0u 0

oi 1 1 4 0 4 604a U : 4v4 ao) a. i L) > a > u C:

o 6-4 . -- ,40

Ix 04 TO-~CLa 0a ~ . .6 0 >-9

- 60 9 a 0 a.~~~46t a- . 9' . 00 9 '- 4 .

o 0.-'6 0 .9 a 00 6 a .W 9 .

-4~~~.4 .- 4 a 'A~ a

.O a, U

a 6.9 6>6 9., 9

6 6 6:O a44 0L a

~ 6 a30

Oo ffib6

1 ' -

I

in the performance of those tasks. Similarly if a "low" amount of a per-ceptual/psychomothr ability is required for the performance of assignedtasks within an AFSC, it would follow that individual differences could beexpected to be muted in the performance of those tasks.

The patterns of perceptual/psychomotor abilities emerging in Ta-ble 4-14 complement the patterns of perceptual/psychomotor abilities al-ready reported in Table 4-9. The inferences made concerning the patternsin connection with the judgments of the amount of 13 perceptual/psycho-motor abilities can be given an additional weight in personnel classifica-tion assignments to an aptitude area. For, as performance quality vari-

ability as a function of the 13 perceptual/psychomotor abilities increases,the probability of objective achievement within an AFSC decreases withcorresponding increases in cost of training and retraining. The more ho-

mogeneous the work group in terms of raw talent and operational efficien-

cy, the more productive is that work group toward realizing group aims.Thus, "high" performance quality variability as a function of the 13 per-ceptual/ psychomotor abilities signals a need for reevaluation of classifi-cation assignments. "Low" performance quality variability as a functionof 13 perceptual/psychomotor abilities confirms that individual differ-ences in the operational display of an ability is not detrimental to the ef-forts of the work group toward the achievement of designated goals.

Comparison of the grand means for the previous amount data andfor the performance quality variability judgments indicates the same fourhighest and four lowest abilities to be involved in both cases.

Task Information- -Amount

The sampled airmen judged the amount of the 13 perceptual/psy-chomotor abilities required in the efficient performance of each of 60tasks, as previously described. The judgments were reduced to a task

mean for each task in terms of the 13 abilities. Accordingly, 780 (60 x13) task means were calculated for each of the 35 AFSCs.

To ascertain whether the 60 tasks in an AFSC differed from oneanother in terms of the 13 perceptual/ psychomotor abilities required in

the efficient performance of a task, each task mean was compared withthe combined mean for supervisors and subordinates for that ability'samount in an AFSC. If a task mean was one-half SD above the combinedmean for supervisors and subordinates for an ability, that task was

judged to require a relatively high amount of the ability in the cfficient

performance of the task. If a task mean was one-half SD below the com-

bined mean for supervisors and subordinates, that task was judged to re-

quire a relatively low amount of that ability in the efficient performance

of that task. This procedure compares each task to every other task on

131

an individual ability basis. The frequencies of task means deviating one-half SD unit above (high) and below (low) the reported combined meansfor supervisors and subordinates for each of the 13 perceptual/psycho-motor abilities in each AFSC according to the four aptitude areas are re-ported in Tables 4-15 through 4-18.

It can be noted, in general, that 2,629 out of 27, 300 (10%) taskmeans deviated one-half SD unit above the perceptual/psychomotor abil-ity mean over the 35 AFSCs and that 1,772 out of 27,300 (6 %) task meansdeviated one-half unit below the perceptual/psychomotor ability meansover the 35 AFSCs. The two tasks in each ability which were highestare identified in Table 4-19.

Aptitude areas differed from one another in terms of the fiequen-cy of relatively "high" and "low" tasks identified for each of the 13 abil-ities. In the administrative aptitude area, only 468 out of 7, 021 (7%)task means deviated one-half SD unit above the selected criterion meansand 297 out of 7, 020 (4%) task means deviated below the selected crite-rion means. In the mechanical aptitude area, 732 out of 6,240 (12%)task means deviated above the selected criterion means and 508 (8%)task means deviated below the selected criterion means. Similarly, inthe electronics aptitude area, 784 out of 7,020 (11%) task means devi-ated above the selected criterion means and 599 out of 7,020 (9%) taskmeans deviated below the selected criterion means. In the general ap-titude area, 644 out of 7,020 (9%) task means deviated above and 368 (5%)task means deviated below the selected criterion means. Sampled air-men in the four aptitude areas, therefore, expressed differential judg-ments concerning the amount of the 13 perceptual/ psychomotor abilitiesrequired in the efficient performance of the 60 tasks in an aptitude area.It can be surmised that enlisted personnel are sensitive to the require-ments of efficient job performance and that different aptitude areas aredistinguished by different patterns of perceptual/psychomotor abilitiesfor the efficient performance of assigned tasks.

Moreover, scrutiny of eachAFSC confirms the emergence of dif-ferential patterns of "high" and "low" deviating task means with respectto the 13 perceptual/psychomotor abilities. It is clear from Tables 4-15 through 4-18 that different tasks, as would be expected, require dif-ferent amounts of a perceptual/ psychomotor ability.

The relative significance of a perceptual/psychomotor ability inthe efficient performance of the tasks of an AFSC can be abstracted fromTables 4-15 through 4-18. Thus, for example, "high" and "low" devi-ations ocLur very frequently in connection with finger dexterity, manualdexterity, and control precision but considerably less frequently in con-nection with rate control, depth perception, divided attention, and kin-edthetic memur. Weights, therefore, could be assigned to AFI'SC tasks

132

17 'D

-Cd' - ' 3 0

o U) N % N .- - -

cui

-d .0N - '

E-4 CO-

.2 -, u 30 -Z

o4 0 .- V)~

- N N,

-4 1

0 U--

C.)

44> .CU --

0 U2

I I N '

-d 0 0

0N

~Cd

r z4Cfj ul-

- ~ N 0 N 3133*

4. Cd : - N - -

-, 41 ~ ,~

Cd

co " .- 3 - - I ID I - - 1 0

14 Q)

4' cz,

a)4

$.4~

Cfl4

0- ti A' N I I

F-. ..-

~~~1 >4 2 - - i a i N - I '

.- tj A0

F-'I 7

134

.0 -, 0 1i 0 0 O .n - .

0 C- 100I . N ~ -

_4 Cd

Q) 0

0.

U) (L

>'~C>

- 0)

z. 0 E 0 ~ 0 '

-~~- 1 7 0 .1 I

0U H -El1.. N C 41 .

C) .7

4U (L , - $1 3

UC - I I

go -, 0 I 7- 0

-n w 00 e0 1 0 0

-4 0 -C C NNs ' 0 -

44 Cd

Or -u C

IIs,

- i c0 en 41N~ - C' I N

j CCmU

115

o~~ ~ c - %V- C 'N 3

-~ IN I I I I I

.4I C-

0l) r-' I N > I N I ~

-- N

-d .3 cc 7Iz 3

0) C.)

Q) tIc cc

CC) - I4 I I I

- ' - IcIcc

r-: ) o I C - I c'

cU 0

o >: Ic c C

o ..- -;

4 (~) c-d Cd N 3 Nc

<Ucc

d) CC',N -3 .

c~cU

Cd'- " . 3 - - *' - N C

C.I X CU)u

3: N t' - .3 : <

0)>~~ - .- -- C

CC -p - : N : ZS

v c) V - N z cc

-136

Table 4-19

Two Tasks Which Were Highest on EachPerceptual! Psychomotor Ability

Ability Task (Mean Rating*. Specialty)

Finger Draw symbols, emblems or pictures on graphics or mastersDexterity (6.53*, Graphics)

Letter graphics products such as charts, posters, orcertificates using freehand pen techniques (6.44, Graphics)

Manual Solder, splice or replace wiring or connectors inDexterity instrument systems (6.00, Avionics Instrument Systems)

Assemble or wire radio or auxiliary equipment componentsfor installation (6.10, Ground Radio Equipment Repair)

Control Operate standard gasoline or electric powered forkliftsPrecision (6.23, Munitions Maintenance)

Operate munitions transport trucks or truck-tractors(6.19, Munitions Maintenance)

Rate operate standard gasoline or electric powered forkliftsControl (6.08, Munitions Maintenance)

Operate munitions transport trucks or truck-tractors(6.00, Munitions Maintenance)

Visual Calibrate or adjust heads-up display (6.33 AvionicsMemory Instrument Systems)

Compare climatological factors with weather systems(5.89, Weather Forecaster)

Visual Remove and inspect engine bearings (6.13, Jet EngineSpeed & Mechanic)Accuracy Locate meteorological features on charts (5.90, Weather

Forecaster)

Position Perform preflight or postf light inspections on staticMemory discharges (6.00, Radio Operator)

Prepare aircraft for engine removal or installation(5.96, Jet Engine Mechanic)

Auditory Operate rotating antenna equipment for maximum signalDiscrim- strength (6.13, Radio Operator)ination Make receiver changes or adjustments to reduce inter-

ference (6.13, Radio Operator)

137

7"7

Table 4-19 (cont.)

Auditory Report interference caused by jamming (5.87, RadioMemory Operator)

Determine type of interference (5.53, Radio Operator)

Clerical Perform graphical or statistical analysis of technicalPerception studies (6.13, Weather Forecaster)

Perform pure system analysis (5.89, Weather Forecaster)

Depth Establish landing sequences (5.70, Air Traffic ControlPerception Operator)

Hold arriving VFR aircraft at visual fixes (5.19, AirTraffic Control Operator)

Divided Establish landing sequences (6.26, Air Traffic ControlAttention Operator)

Test personnel under operational conditions (6.11, AirTraffic Control Operator)

Kinesthetic Load film onto reels (4.92, Still Photographic)Memory Remove and install components within fuel cells (4.70,

Aircraft Fuel Systems Mechanic)

1 3

in terms of the amount of the ability judged to be required for efficientperformance of those tasks. That is to say, perceptual/psychomotorabilities vary in their contributions to job efficiency; attention needs tobe focused on some abilities and withdrawn from others.

Task Information--Performance Quality Variability

To ascertain whether the 60 tasks in an AFSC differed from oneanother in terms of performance quality variability as a function of the13 perceptual/psychomotor abilities, each task mean was compared withthe combined mean for supervisors and subordinates. Accordingly, 780comparisons were made in each AFSC separately. If a task mean wasone-half SD unit above the ability, that task was judged to display a rel-atively "high" performance quality variability as a function of the 13 per-ceptual/ psychomotor abilities. If a task mean was one-half SD unit be-low the ability mean, that task was judged to display a relatively low per-formance quality variability. The frequencies of task means deviatingone-half SD unit above (high) and below (low) the combined means arereported in Tables 4-20 through 4-23.

Of the total, 1,261 (4.6%) task means deviated one-half SD unitabove the ability mean combined for supervisors and subordinates withrespect to performance quality variability and 635 (2%) task means de-viated one-half SD unit below the ability mean with respect to perform-ance quality variability. On the general level, these data suggest thatpertormance quality variability as a function of the 13 perceptual/ psy-chomotor abilities was not perceived as being a significant variable inassessing job efficiency. In general, the findings suggest that the sam-pie airmen believed efficient performance of assigned tasks in their re-spective AFSCs to be fairly uniform as a function of the abilities con-sidered.

More specifically, the four aptitude areas did not differ from oneanother appreciably in terms of the frequency of relatively "high" and"low" tasks identified for each of the 13 abilities. In each of the fouraptitude areas, approximately 1 percent of the task means deviated one-halt SD unit above the selected criterion mean. In the mechanical andthe electronics aptitude areas, approximately 1 per ent of the taskmeans deviated one-half unit below the selected criterion mean, andless than 1 percent so deviated in both the administrative and the gen-eral aptitude areas.

The relative significance of the pereptual/ psychomotor analysisrelative to performance quality variability c.an he abstracted from Tables4-20 ihrough 4-23. Frequencies of such deviations also varied onlyslighily between and within A1'S('s.

139J

4

- 3 . -d~

r- !'-1-

-~ N 'D

- C.d

=a a) '3 I I N d

a ~ o :

04 a- 4' U

CA)>4 I I N I I I N I '

woco -4 >4 - '3 N - -

0 0

0) >0- Cd 04 - 0 N N

.0 -~T

c14

.1 .- - -

=S -a

. -4 0

ci

~0 C

.-. Iz) c ZClI

.16 7 - N l-4 U) -,- -l0-

olA

Cd .74-7 l - N N

o S 4. - ~ a~~ - .

CUa

- - I -7 -

co -

o e7

tic -a)

SZ ~ . 7 ~ .1 <

0 I .4

- - - -

od'- ' d

14

i -

C) 0

i- I N N -: -- N .

Cli

4.,

S' o -

C)) L I 9 I

0 cU

co - 3

0o u

clo

C13.

I I -C

- ~ ~ ~ ~ .0 00 I '

w 0

2 ~ =N ~ . ,-. IN143

Fl4

Agi'eement Between Amount and Perfor mance

As noted earlier, thle amiount of 13 perceptual/ps ' chomotor abil-ities in the efficient performance of 60 tasks was judged by' supervisorsand subordinates. Similarly, supervisors and subordinates judged pet -

formance quality variability as a function of' the 13 abilities, The twoseries of'preta! scooo ability uneans allow for a direct comn-parison. Accordingly, thle two sets of' means were separately rankedfrom low to high by Ai'SC. These rankings were separately graphed inlF~'igures 4-1 and 4-2 along with thle boundaries of' one-half SD unit aboveand below thle mean.

Thle lowest and thle highest five ranked ability amount means werecompared with the lowest and the highest five ranked performance qual -

ity variability means, respectively. It was merely sufficient that, with-in an AI"SC, the niean of a per-ceptutal sxctliomiotoi' ability appeared inthle appropriate set without regard to hier-archical position within the set.Thle frequenlcy of matches within thle "low"' and ''highi sets was reducedto a percentagye. The per-centage was designated as a fiercentage agr~ee-

ruent index. The perce ntage agreement indices are reported in 'alolu4-24 according to thle four aptitude areas.

It can be imimediately noted that 30 out of :35 (861") of thle per -centage agreemient indices are 90 anid 10)0. The lowest index was obtain-edi for Riadio operator. 'fhese diata suggest thiat orresponden,,e exi-stsbetween thle amiount of a skill required in efficient purf'ormarhe of a taskand thle perceived performance qualit 'y va tiabil it, u ssocLat ed xv ith isu clperifor manc e. 1If it (-an be as sumIied tha t. i ll 1et esittiea110 11 nt Of XI t

ability- required foi- etficienit peicr m la the canl he tos udo meanl ill-creased criticality for, task pe rforrIiaarrc e, thelr it follows that sensitl\ it

to individual differencie in effii-ieit. pelt otrmarce cild hie '.Xpeiwled tO) k'0

var v wit h 1hat crtit ical level. The highecr tlic p err e itage ag'reemlent in -

d ex, thle mlore (105e 'v related wvould be the( allmult of skill arid pIerf,krtil

alice qural ity va riabil itx , i. e. , one could riot prude t0% bye dissciat ed

f romf thle Other-.

Moreover, thle Slopes of,1c th corirparahle sets- (if ki Il\es pa'ah'I

each other clIoselyv. The I requerics ol' hioth sets of rai ike-d means arl-

rep orted in TFable 4-25. ThIUS, rate conr ol , ulit orv di si I iii itat ionl,audiitorv trnr'trov, arid] kinesthetic, memvor\ chiatarIk ~'Isticalv. \xx ie a-

nIonLg the lowest five tirikecl means inl the two sori(5 of' jildi~2tllls.

Fintger d externt. , tria trual dIext e r if.., xisuta I te1110 ry, vis ual speed t tid

airil t'ar' alldi IMoSi ol n ervroiv eli tart; e it ira!11. \ C 1'( cii 1a MI t l(rr 1h IL j-I(.., five, r'arkel Ileatis ini the twok si ries, ot udmt .\c Ai-ndrIll) ll

reltie tthi~r~iri ii hese skilIs Is furtrilfl estIalishied ill tire effi e'

1-14

AhRCFAF7 UACMAS-7ER NTELL!GEC- OPEAY101S A.140

4MON T IMAGERY INERPFE7AMCNAMCLNT

. A0

CLTP 0&

5... A NA 5. AO~t

A AC OA

6 A

t .. cLP

Ic :)'pJ"

A A

A. A

C-

X tA

aO m Xto

FDP'~C STine1DxertIPL Potion Memor

AMOUNotro rcso AM=AdtoyMmr

IA LPA

VM Xu eA

A = is

c-PP

L3AA

C A 4

FD Fige DetrtyP PstonMmr

CP = Coto rcso M uioyMmr

AM = ViulMmr A= iie teto

VA A AiulSedadAcrc M=KnshtcMmr

CLP = Clerical Perception

Figure 4 1. Mean and + .5 standard deviation for each abilityfor judgme--nt of amount.

ADS lWAPPI CWT11iOL O m1m TECHWI0N RAW0 OPERATORAMOUNT AMOUNT

m I

I~ x m

x A

GROUND ~URADIO EOUIPMENT REPAIR MIm tm •X if SSILE u&ONEL.ECTRONI l¢ OUIPMENT

AMOUNTa A K K

AIOUN C ENTUMT r'TSEAIST MiSIt.RE' E.CTRtdC EIP. 'MNTl*t.

AMOUNT S1P C AMOUN

P.0

x a a

xx

D 7

oox

x x

Figure 4-1 Continued.

146

ETGINE MECHANICAIRCRAFT FUEL SYSTIS MECHANIC AMOUNT

AMOUNT

4 Pt

3L CA a, *

" ".

LGM 25 MISSILE MECIHAINIC AMOUNT

AMOUNT

4.

0 ,, PAP

4 ?J "l

3 op

000

: - AP MEC PANI * M SLW A LII

* MUN S P

- I P•* 1 0

147

* N

ELECTRICIAN ELECTRICAL POWER PROOUCTION

AMOUNT *O AMOUNT

A 3-

CL 0 1~L0 CIN

* 0*** No N'

A A x

V a

F CARPENTRY AND MASONRY ,PCIA.ST FIRE PROTECTION

AMOUNT AMOUNT

0 0

I I

N

NC N

00n oc

FV E I L P R T R C S A C E I A S N E N AIR C AR G S P IA CIS

DP CLP AM0 A C 0N. x0

Au A

FIg u• 4 - C o nt• xe d

1 4ox 9

S VEHICLE OER TOR/OISPATCHER AIR PASSENG R AN AIR CARGO SPCIALIST

AMOUNT AMOUNT

D, IS ILI

Am p DA

x

igure 4-1 Continued

148

L SUPPL.Y SEVCSFUELS SPKEIAUTJ~AMOUNT AMOUNT

N Ac

x N A cA

CLx .p€ MC 'P D&

AMOUNT

CLP

x N

A N C

A4 A' N' fCI NIANTERY MANAGEMENT IRM N .MI IA I"

AMND U T NYNTN AMOUNT4 N o

Ccp cLA x (WA

OF N

20

MAiEEN F Aontinued

N N

149,

CCA MOM TI N-NInoC4, N I NAN*

NI. .. . vs NCtCC' * I Co. OA * N .NM N

* N OP NC A

F qurs 4-I Coxntgnued

r 149J

LEGAL SERVICES PERSONNEL

AMOUNT ' AMOUNT "p

.SG

- 4I "

.c.

a. K op

K K,

9 AA4JN ASP AMOUNK

.0 K U

O A a

AMOUNT KS KLPP AOD

2 A. AD It x

I ol

AK K

K K

II

INENTAL NDTPREVETIVE DENISTSly E4ICIALN[VI

AMOUNTVM €IAMOUNUN

.0!

2 KS A K K K K

I xi

DEOFNTAL AND pREtVIVI DENTISTRY TIrCONICIAN

AMOUNT

SS

0,

K K

Figuru 4-I, Continued.

150

F W

AIRCRAFT LOADMASTER INTELLiGENCE OPERATIONS 4ND iMAGERY

PERFORMANCE QIJALITY VARIAILITY INTERPRETATION"1 s L PERFORMANCE 0UAUTY VARIABILITY

At

A L

CLP -K OA AlA PM FO

AX A M M

CX X X X

X 21 X

CP

X X X - x - . . . •w A

GRAPHICS STILL PHOTOGRAPHIC

PEFORMANCE QUALITY VARIABILITY PERFORMANCE QUAITY vAMA.IUY

4 3 L V A

A ALPCLi P * x UI Al A" A* .

A A4 CP X

X4 " W, DA

4S* A AA

AA A * A

4-ra * A . A A

A[ 'r

IEATHER CORCASTER A)RPORTAIR OPERATIONS

PERFORMANCE UALIT A 3A P PERFORMANCE DUAL1TY VARIABILTY-VARIABIUTY •

xX CL

opA M SA II

2 C, * K (MAA

' • , A .. A A A 4• M ,* • , 0* x

p RC A A

FD = Finger Dexterity PM = Position MemoryAD = Auditory Discrimination MD = Manual DexterityCP = Control Precision AM = Auditory Memory

RC = Rate Control DP = Depth PerceptionVM = Visual Memory DA = Divided AttentionVSA = Visual Speed and Accuracy KM = Kinesthetic Memory

CLP = Clerical PerceptionFigure 4-2. Mean and + .5 standard deviation for each ability

for judgments of quality and variability.

151

S AIR TRAFMI CON4TROL OPERATOR TECHNICIANC RADIO OPERATOR5 PERFORMANCE QUALITY VARIA9BILITY gPERFORMANCE QUALITY VARIABILITY

0A

41 4C

CA AD As, AL

A- K

GROUNO RADIO EQUIPMENT REPAIR MISSILE ELECTRONIC EQUIPMENT SPECIALISTPERFORMANsCE QUALITY VARIABILIT PERFORANCE QUALITY VARIABI1LITY

A,.. 4

N No AI

N t.1N C N A

N = NCA N0

D.~ CA

Ni A 0 CINON

Fiur 4- CoNntinued

I C N52

I--GaN

AIRCRAFT FUEL SYSTEMS MECHANIC JET ENGINE MECHANIC

PERFORMAJNCE QUALITY VARIABILITY PERFORMANCE QUAUTY VARIABILJTY

F.

* I-

' x x

LGM 25 MISSILE MECHANIC MUNITIONS MAINTENANCE

PERFORMANCE QUALITY VARIABILITY PERFORMANCE QUALITY IARIABILITv

( ~~~ m: ..D x •

AO "

v. ft x

ADLP I C

2M IS MIHA*C MUIIN MANTNAC

PERFORMANCE QUALITY VANIABlUTY PERFORMANvE QUJALITY VARIAI*tLITV

VS VI

I S

• - WEPN MEHAI - M oSI[ ACLII

I PEF*MN ULT * iIWPEFRAC QUAITYv* Itl

Figur 4-2 COnfInued

15F3

ELECTRICIAN ELECTRICAL POWER PRODUCTIONPERGRMU ICE QUALITY VARIAIIlULTV PERFORMANCE OUAUTY VARIABILUTY

* A

cp

AO Cl I• Da ac

NA N A

x x x x

PERFORNC QULT VAIBLT 5 PROMNC A ULT AIBLT

x vu o v"

4 x 4 x

xx

o V I 0 AI A N AND A

CAIJOThV AND I SMOiS? FIRE O*TE£CTON

PERFORMANC ITy VA U PERRMANCE OUALITY VARIABILITY

N eV.. A N C A A

Fiur 4-2 Continued

3N aC A N

4 * A AA N

N N

N 4VE HICLEF OPERATOR/ISMTCIE AIR PASSErNGErM AND AIR CARG SPEOIALIST

PErRFORMANCE QUALITY VARIABIUTY' PERl IMANE ONAUTY VAJRIABILITY

V14CLP NA VN

AM AN AI N

N , Se I

N IN

PIgurm 4-2 CamtIuwd

154

"4

SUPPLY SERVICES FUELS SPECLAUST

PERFORMANCE QUALITY VARIABILITY PERFORMANCE QUAU VARIABIUTY

WEQULTYVRIABILIT

INC. x

me cp c ,

0 7

K m

K £

0 4 PE CEo

_

4 ItOENTOW MANAGEMENT, MATERIEL FACIUTIES PROCUREMENT SPECIALST

AD SUPPLY SYSTEMS PERFORMANCE QUAUTY VARIABILIT Y

PERFORMANCE QUAUTY VARIANUTY

CLA U .

OA 1op A

40 6" op x ¢

¢

K 9 K K K

CHAPEL MANAWT AIRMAN AQ&INISTRATION

pEpFORMANC QUAUTYVARIABIUTY PERFORMANCE QUAUTY VARIABILITY

KK KCL

K K c'

K P K K

K p n K *e Ko KKop

F.qurs 4-2 COKninuld

155

LEGAL. SERVE PESNEPERF MiANCE OAUTA Y VARIABILITY I PERFORMANCE QUALITY VARIABILITY

.00.0 x

x K

c x ko

" K *

INAORMAOTIONSP A0ST MEDIC.AL SPERViENIVIE \TE I

5IRORAC PERFO VA R MANCELIT r PAR(ABI MAN CI AIYV AI

coo

Figur 4- 0ofne

15

PM ,p~ ~ Sh NO

K 0I

°, .. O'

i~f

Table 4-24

Agreement Indices Between Means of Judgments of Amountand of Performance Quality Variability

AFSC Agreement AFSC Agreement

Mechanical

Munitions Maintenance 100 Weather Forecaster 100

Weapons Mechanic 100 Information Specialist 90

Air Passenger and Air Inventory Management,Cargo Specialist 80 Materiel Facilities

Aircraft Loadmaster 90 and Supply Systems 90LGM 25 Missile Mechanic 90 Medical Services 90

Vehicle Operator/ Dental and PreventiveDispatcher 100 Denistry Technician 90

Fuels Specialist 100 Fire Protection 90

Carpentry and Masonry Graphics 80Specialist 90 Still Photographic 100

Administrative Electronics

Procurement Specialist 100 Missile Electronic

Chapel Management 100 Equipment Specialist 100

Legal Services 90 Missile Facilities 80

Radio Operator 60 Aircraft Electrical Systems

Personnel 100 Specialist 90

Air Traffic Control Avionic Instrument Systems

Operator Technician 90 Specialist 90

Airman Administration 90 Ground Radio EquipmentRepair 80

Supply Services 100 Electrician 90

Airport Air Operations 100 Electrical Power

Production 90

General Aircraft Fuel SystemsMechanic 100

Intelligence Operations Jet Engine Mechanic 90and Imagery Interpre-tation 90

'Fable 4-25

Frequency of Occurrence of Ability Means for Amountand of Performance Quality Variability

in High and Low Categories

Amount Quality

Ability Low High Low High

Finger Dexterity - 31 1 31

Manual Dexterity 2 29 1 28

Control Precision 3 2 7 4

Rate Control 23 1 23 -

Visual Memory - 35 - 33

Visual Speed and Accuracy - 32 - 30

Precision Memory - 28 - 32

Auditory Discrimination 31 1 32 -

Auditory Memory 32 - 34 -

Clerical Perception 7 9 6 9

Depth Perception 26 1 29 1

Divided Attention 13 4 14 3

Kinesthetic Memory 27 - 28 1

158

performance of assigned AFSC tasks. It could be that these skills aremore or less necessary for effective duty performance and should beseriously attended to in career assignments.

Pearson coefficients of correlation were also computed betweenability means and the performance quality variability means across the35 AFSCs. The correlation coefficient for each AFSC is reported inTable 4-26 according to the four aptitude areas.

It can be immediately noted that 32 (91%) of the obtained coeffi-cients exceed . 900. The lowest correlation was obtained for Radio Op-erator. The correlation coefficients for Air Passenger and Air CargoSpecialist and for Ground Radio Equipment Repairman were . 794 and.820, respectively. The median correlation coefficient was . 971. Ahigh correspondence was again demonstrated, therefore, between theability means and performance quality variability means. For all prac-tical purposes, the two series of judgments were communicating essen-tially the same information concerning the role of the 13 perceptual/psy-chomotor abilities in the efficient performance of the tasks of an AFSC.Differential perceptions concerning the amount of a skill and perform-ance quality covaried, as a rule. Common factors seemed to have beenoperating in influencing the sampled airmen in rendering the two seriesof judgments.

Supervisor-Subordinate Agreement

Product moment correlation coefficients were calculated betweenthe mean data for both the amount and performance quality variabilityratings made by the supervisors and the subordinates within each careerfield. The resulting correlation coefficients are presented in Table 4-27.

With the exception of two AFSCs (Air Passenger and Cargo Spe-cialist and Radio Operator), the correlation between supervisors andsubordinates was very high. An examination of the individual judgmentsfor the two career fields with low supervisor-subordinate correlationsrevealed that two supervisors in each group were divergent from theothers. The effects of the two divergent sets of judgments were power-ful enough, with the relative sample sizes, to depress the resultant cor-relation coefficients.

The correlations for the career lfields, in 'Fable 4-27, were con-verted to z' coefficients and averaged. The average correlations, acrossthe 35 career fields, were .94 for amount judgments and .90 for per-formance quality judgments. it is apparent from correlations as highas those obtained that supervisors and] subordinates, within a gi\cn

159

Table 4-26

Correlation (r) Between Amount of 13 Perceptual/ PsychomotorSkills and Judgments of Performance Quality Variability

AFSC r AFSC r

Mechanical

Munitions Maintenance 0.978 Weather Forecaster 0.992Weapons Mechanic 0.974 Information Specialist 0.972Air Passenger and Air Inventory Management,

Cargo Specialist 0.794 Materiel Facilities

Aircraft Loadmaster 0.946 and Supply Systems 0.985LGM 25 Missile Mechanic 0.968 Medical Services 0.954

Vehicle Operator/ Dental and Preventive

Dispatcher 0.949 Dentistry Technician 0.980

Fuels Specialist 0.983 Fire Protection 0.974

Carpentry and Masonry Graphics 0.987Specialist 0.967 Still Photographic 0.986

Administrative Electronics

Procurement Specialist 0.993 Missile Electronic

Chapel Management 0.953 Equipment Specialist 0.959

Legal Services 0.994 Missile Facilities 0.931

Radio Operator 0.680 Aircraft Electrical

Personnel 0.993 Systems Specialist 0.953

Air Traffic Control Avionic InstrumentOperator Technician 0.976 Systems Specialist 0.930

Airman Administration 0.960 Ground Radio Equipment

Supply Services 0.963 Repair 0.820

Airport Air Operations 0.959 Electrician 0.971

Electrical PowerProduction 0.970

General Aircraft Fuel Systems

Mechanic 0.975Intelligence Operations Jet Engine Mechanic 0.975

and Imagery Interpre-

tation 0.989

I 6U

Table 4-27

Correlation Between Supervisor and Subordinate Ratings for

Amount (A) and Performance Quality Variability (PQV) Judgments

Correlation

FSC A PoV

Mechanical

Munitions Maintenance .94 .90Aeapons Mechanic .98 .88

Air Passenger and Air Cargo Specializt .29 .76

Aircraft Loadmaster .93 .89LOM 25 Missile Mechanic .98 .91

Vehicle Operator/Dispatcher .95 76Poels Specialist .95 .37

-ir)entry and Masonry Specialist .85 .5

Ad:iistrative

?rocur~ment Specialiat .97 .98

Chapel Management .98 .96Legal Services .94 .92

Radio Operator .63 .50

Personnel .95 .94

Air Traffic Control Dperator Technician .89 92Airman Administration .97 .92

Supply Services .32 .77

Airport Air Operations .S4 .70

General

Intelligence Operations and Imag ry Interpretation .95 .97

Weather Forecaster .97 .95

Information Specialist .90 .77Inventory Management Materiel Facilities and Supply Systems .94 .97

Medical Services .90 .82Dental and Preventive Dentistry Tecimnician .89 .16Fire Protection .96 .82Graphics .96 .91Still Photographic .96 .98

Electronics

Missile Electronic Equipment Soecialist .94 .35Missile Facilities .97 .33Aircraft Electrical Systems Specialist .91 .83

Avionic Instrument Systems Specialist .99 .95Ground Radio Equipment Repair .33 .86Electrician .94 . 7Electrical Power Production .96 .94Air:raf: Fuel Systems Ye.hanic .93 .96

Jet Engine Me:hanic .) .76

Average .94 .90

161

-- Nowm

career field, may be regarded as representative of a single Air Force

population of job incumbents for the purpose of job analytic judgmentssuch as those made in the present survey.

Profile Similarity

An index of profile similarity was computed to compare the pro-files of the 35 AFSCs. The amount and the performance quality vari-ability profiles were considered separately. The index of profile sim-ilarity employed was the D statistic (Osgood and Suci, 1952). The equa-tion for computing the D statistic is:

D = Id 2

whe re:

d = difference between the score on an abilityfor one specialty and the score on thesame ability for another specialty.

The equation considers both the shape and the elevation of theprofiles compared and is superior to the product moment correlationin this regard because the product moment correlation considers onlyshape. The D statistic was computed among all AFSCs (35 x 34/2) foramount and for performance quality variability. The profiles were cat-egorized as essentially similar or essentially dissimilar. The essen-tially similar criterion was a D score of 2. 52 (10%) of the maximumrange of the D statistic or less. The percentage of the specialties whichmet the "essentially similar with the other AFSCs criterion" was com-puted. The results are shown in Table 4-28. For amount, the WeatherForecaster AFSC never met the criterion and the Medical Services Spe-cialist AFSC met the criterion for 76 percent of the comparisons. Ac-cordingly, the Weather Forecaster AFSC is more independent than theMedical Services Specialist AFSC. The median percentage in Table 4-28 is 50 for amount and 65 for performance quality variability. There- mne evidence for partial, but not high, independency among the

r. lat ions Among Abilities and Factor Analysis--Amount

.. ,,ari ratings of the amount of each ability (across 60 tasks), 'J v wre intercorrelated to yield a 13 by 13 matrix. The

x )f intercorrelations is shown in Table 4-29. A sum-, ,uorment correlation coefficients follows:

162

Table 4-28

Percentage of D Scores in Each AFSCWithin 10 Percent of the D Range (2. 52) or Less

Specialty Amount % PQV %*

Aircraft Loadmaster 68 82Intelligence Operations and Imagery Interpretation 41 44Graphics 35 62Still Photographic 65 76Weather Forecaster 0 0Airport Air Operations 18 29Air Traffic Control Operator Technician 0 24Radio Operator 6 65Ground Radio Equipment Repair 50 62Missile Electronic Equipment Specialist 59 74Avionic Instrument Systems Specialist 9 35Aircraft Electrical Systems Specialist 32 26Aircraft Fuel Systems Mechanic 62 76Jet Engine Mechanic 59 79LGM 25 Missile Mechanic 53 62Munitions Maintenance 65 79Weapons Mechanic 32 65Missile Facilities 53 68Electrician 53 68Electrical Power Production 50 65Carpentry and Mascnry Specialist 53 59Fire Protection 38 26Vehicle Operator/Dispatcher 32 41Air Passenger and Air Cargo Specialist 32 18Supply Services 68 74Fuels Specialist 56 76Inventory Management, Materiel Facilities and

Supply Systems 38 56Procurement Specialist 18 29Chapel Management 53 74Airman Administration 53 74Legal Services 26 59Personnel 18 56Information Specialist 38 68Medical Services 76 85Dental and Preventive Dentistry Technician 59 71

*PQV =Performance Quality Variability

163

.-. 1- 00 O% N 0% .- 4 -- 4 D 00 OD 0

o) C 0C0 0 0) 0 0 0 0) 0 0 -

N1 0 Vf) 0D M Lf) m- 0 0 -11 00 0

o00000 0 0D 0 0 0 -

C)0 0) 0 0 0 0) 0) 0 0 0 -

Cd

;o co 0n -. ? m' :r . 0 - -

CUCD No 0D N O . 0 CO 0

-- 4 '.D %D 1-1 cn Lfn 0% 0 1

0- 04 10 10 04 0 0 CD I-

(9C N; COC;N7-

0 0 r- 0 0 0 00 0 I

2J= ~ Lr 0 0 0 0 0

En -'r Ir 't o C I

HC CU0 D 0 0 0 4$

ON 0 0

0

0 0 '-

CCU

1-G

N %

_>.50< .99 35 45

>.30< .49 20 26

>. 10< .29 9 12

<. 10 14 18

These data suggest that a moderate degree of uniqueness existsamong the ratings of the perceptual/psychomotor abilities for Air Forcejob tasks.

The intercorrelation matrix was factor analyzed by the principalcomponents method with orthogonal rotation employing the SPSS PA2package (Nie et al, 1975). Three factors emerged employing the eigen-value of 1. 00 or greater as the criterion for accepting a factor. Theeigenvalues and respective cumulative proportion of variance were:

Cumulative %

Eigenvalue of Variance

Factor 1 6.71 51.6

Factor II 2.33 69.5

Factor 111 1.86 83.8

The two abilities loading highest on each factor along with therespective loadings were:

Abilities LoadinglIiglhest Loading

Factor I Auditor\ Discriinination .96Auditor5. Menory .93

Factor II Visual Speed and Accuracy .86Visual Memory .85

Factor II Manual Dexterity .92Fingezr Dexte'ily .69

Attempts to disthi b this solution h rough oblique, ital.hir than or -thogonal rotation produced no nar mked effect. The results of lie facloratnalysis suggest, as was probably to be anticipated, a visual f& ' or, ali

auditor\ fad(tom, aird U mauIIal fa(ator.

I1i5

Reliability Analysis

It is well known that the reliability of ratings can be increased itmultiple judges' ratings are averaged to yield a single measure. B'pooling different judges' ratings into a composite measure, tile relia-bility canbe increased substantially, even with only a moderate increasein the number of judges. This point has been made explicit b\ Winer(1962) and by Cronbach, Gleser, Nanda, and Rajaratnam (1972). He-cently, it has also beendemonstrated that the averaged ratings of morejudges are in fact more valid (Horowitz, Inouye, and Siegelman, 1979).Other things being equal, a measure that is less reliable cannot corre-late as highly with an external criterion. As one icreases the reliabil-ity of a judgment by averaging the ratings of an increasing number ofjudges, one can expect the resulting measure to correlate more highlywith an external criterion.

For four of the AFSCs (one randomly selected from each aptitudearea), the data of eight randomly selected supervisors and eight ran-domly selected subordinates were analyzed to determine inlerrater re-liability. The analyses included four abilities and 15 tasks. Eight sep-arate variance analyses were completed--four for subordinates and fourfor supervisory raters. Only amount data were considered. Table 4-30 presents the variance analytic model employed for each of the eightanalyses. If an analysis of variance is performed employing tile modelpresented as Table 4-30, then the reliability of a single judge may beestimated by formula 1. roiom Winier (1962), the f'ormula is adapted asfollows:

MS between treatments - MS cirorr( 1) MS between treatncnts +(n- 1) MS error (1)

where: number of raters (judges) = 8

treatments= 4 abilities, 15 tasks - GO

To provide appropriate values for the formula, belween-treainmnents var-iance was estimated by pooling across abilitics, tasks, and the tasks-by-ability interaction (df- 59). E'ror variance was estimated 1V ' .the iii-teraction involving raters by tasks by abilities (df- 294). By appli'alionof the Spearman-Brown iprediction foi'mnlua, the rliabilit\ of ail I in ml),of judges may also be estinated. l,'or 24 judges the estiimate is give,n h\formula 2.

24(r 1))(2)r(24) - + (24 - 1) (r(1))

where '(i) - reliability (of a si ngl(' judge

166

Table 4-30

ANOVA Model for Reliability Determination

Source of Variation df

Rater (S) (n-1) = 7

Task (A) (p-1) = 14

Ability (B) (q-1) = 3

S x A (n-1)(P-1) = 98

S xB (n-1)(q-1) 21

A xB (p-l)(q-1) = 42

Sx A xB (n-1) (p-1) (q-1) =294

TOTAL npq-l 479

167

Table 4-31 presents reliability estimates for a single judge (11),the average of eight judges (r8), 16 judges (r1(i), and 24 judges (r24).The results suggest that for the data analyzed here, acceptable reliabil-ities require about 24 raters.

Stability

Another view of the reliability of the data can be gained b' exam-ining the agreement between the mean data obtained in the pretest withthat obtained in the major data collection effort for the two specialties,Munitions Maintenance and Fire Protection, which were common to bothstudy aspects. The correlation was obtained between the categoricalscale judgments of the two specialties pretested and those of the equiv-alent judgments from the final data collection. We note that changes inmethods and scaling procedures will serve to reduce the level of suchcoefficients. The following results provide considerable support for tlheemployed technique ts reliability.

Fire Protection (571) rAmount 7 8

Performance Quality Variability .53

Munitions Maintenance (461)

Amount .81Performance Quality Variability . 72

Rater' Error

The use of ho man judg ment (r'ating sCal e 1)l'O( CdUl'Cs) itl 1hp '-sent case t'ests Oil the assLnmp)tion that the holmall observtel' I'tQ esclltsa suitable observational instrument and is capable of achtiev tog p'Cisiolland objectivity. While we have confidence in quatitative lima:i judg-ments, we are also aware of sources of ias ill sIW(h jodginelIts. Ac.otl-d-ingly, the effects of three types of tatinl erl'ot on the data were inxves-tigated. They include rating errors termed b\ (;uilford (1954) as cu-trast, leiiiencx, and halo.

The tenden, y for some ral rs to t'ate tle abil it- t'eqotll-(,d ).\ atask against the rater's own level on the skill is ealled .oIll tI'asl (e'ul.

For example, some tatets who are high oi mna:mial dexietttlv ,\ ImnIhl

systematically tend to rate soc h tasks as ,teiti illg a lo\ allOutllt of tltlsability.

16t;8

Trable 4-31

Retiability of subordinate and Supervisory Rlatingsfor, a Variety of Sam-ple Sizes and AU'SCs

Subordinates No. of Judges

AFSC () (8) (16) (24)

Inventory Management, Materiel Facilitiesand Supply Systems (645X0) .098 .465 .635 .723

Airman Administration (702X0) .269 .746 .855 .898

Weapons Mechanic (462X0) .184 .643 .783 .844

Jet Engine Mechanic (426X2) .249 .726 .841 .888

Supervisors No. of JudgesAFSC (1)- (8) (16) (24)

Inventory Management, Materiel Facilitiesand Supply Systems (645X0) .453 .869 .930 .952

Airman Administration (702X0) .530 .900 .947 .964

Weapons Mechanic (462X0) .152 .589 .741 .811

Jet Engine Mechanic (426X2) .179 .636 .777 .840

Leniency error occurs when raters systematicaltly judge sometasks or abilities to be high or low because they have some interest inthe abilities or tasks. For example, there are "hard" raters and "easy"raters. Some raters may "lean over backwards" to avoid rating anability as more important than it really is.

Halo error would occur in the present context when a judge, whorates a given task as high or low on a given ability tends to carry thisjudgment over to other abilities. One result of the halo effect is toforce the ratings in the direction of a general impression.

The same data set as employed (AFSCs and raters) for the priorreliability analysis was employed in the analysis of rater error.

Eight separate variance analyses were completed, four for sub-ordinates and four for supervisory raters, to define error sources.From avariance summary, it is possible to estimate the strength of as-sociation between the independent and dependent variables. This strengthof association statistic was employed in the present case as a method forinterpreting sources of variance in terms of the proportionate contributionof each source or order of importance (Kirk, 1968, pp. 126, 134, 198).Table 4-32 presents the formulas employed for estimating the strengthof association (Kirk, 1968). Using the triple interaction variance as anestimate of residual variance, conclusions may be obtained with respectto main effects and first order interactions. Variance components are $important in the interpretation of results since even trivial associationamong means may achieve statistical significance if the sample is suf-ficiently large. An association measure states whether or not effectsare large or small independently of the statistical significance of Fratios.

Tables 4-33 and 4-34 present association values (w 2 ) indicatingthe proportion of variance in the dependent variable accounted for byeach of the independent variables. Each dependent variable is a differ-ent specialty. Mean association values across four AFSCs appear inthe last column of each of the tables. The last row of each of the tablesindicates the total proportion of variance accounted for in each dependentvariable. Table 4-33 includes the summary of the analysis of subordi-nate's data. Table 4-34 displays the results from the supervisor anal-ysis.

Association values due to raters indicating leniency error waslow to moderate across all analyses performed. Contrast bias evidencedin the rater-by-ability interactions was also low to moderate in sevenof the eight analyses performed. ltalo bias, evidenced by variance illthe rater-by-task interaction, proved quite low for seven of the eight anal -yses.

170

Table 4-32

Formulas Employed for Estimating Strength of Association (w 2 )Between Independent and Dependent Variables

(S) Rater: w = SSS - (n-1) MSSABSStotal + MSSAB

w SSA - (p-i) MSSAB

(A) Task: XI SStotal + MSSAB

2 SSB - (q-1) MSSAB(B) Ability: W2 XIB SStotal + MSSAB

S x A: W2 XISA = S5SA - (n-1)(p-1) MSSABSStotal + MSsAB

S x B: = SSSB - (n-1)(q-l) MSSAB

SStotal + MSSAB

A x B: W2XIAB - SSAB (p-l)(q-1) MSSAB

SStotal + MSSAB

Note: - S = RatersA = TasksB = Abilitiesn = 8 Raters

p - 15 Tasksq = 4 Abilities

171

Table 4-33

Strength of Association (w 2 ) BetweenIndependent and Dependent Variables for Subordinate Haters*

AFSC Dependent Variable

Independent Variable 4262 4620 6450 7020 Mean

Rater (S) .214 .186 .262 .320 .2455

Tasks (A) .034 .001 .018 .010 .0157

Abilities (B) .093 .071 .010 .059 .0583

S x A .000 .077 .079 .011 .0417

S x B .172 .297 .383 .305 .2892

A x B .029 .008 .000 .028 .0162

SUM (w2 ) .542 .640 .752 .733 .6666

j

*Negative entries were assigned a zero value

(Kirk, 1968, P. 198)

1-2

Table 4-34

Strength of Association (w 2 ) BetweenIndependent and Dependent Variables for Supervisory Haters*

AFSC Dependent Variable

Independent Variable 4262 4620 6430 7020 Mean

Rater (S) .320 .180 .085 .437 .2555

Task (A) .035 .009 .008 .013 .0163

Abilities (B) .005 .042 .341 .071 .1147

S x A .000 .116 .002 .005 .0307

S x B .174 .211 .092 .273 .1875

A x B .055 .023 .019 .067 .0410

SUM (w2) .589 .581 .547 .866 .6457

*Negative entries were assigned a zero value (Kirk, 1968, P. 198)

173

--------------- -------------------

For the subordinates, contrast bias (which is percentage of var-iance in the rater-by-ability interaction) amounted to 28. 9%; leniencyerror (which is variance in the rater main effect) amounted to 24. 6%.Halo bias (which is variance in the rater-by-task interaction) amountedto only 4. 2%.

When the supervisory ratings are considered (Table 4-34), con-trast bias, measured by the rater-by-ability interaction, amounted to18. 7%; leniency error, measured by the rater main effect, amounted to25. 6%. Halo bias, measured by the rater-by-task interaction, amount-ed to only 3. 1%.

174

V. SUMMARY AND CONCLUSIONS

Summary

In order to derive a perceptual/psychomotor skill taxonomy whichcould provide the basis for analyzing the perceptual/psychomotor require -

ments of variousAFSCs, an extensive literature analysis was conducted.The analysis served to isolate perceptual/psychomotor descriptors whichhave been employed previously by others. These descriptors were re-duced through an empirical analysis which compare.d each descriptorwith a variety of screening criteria. The end result was a perceptual/psychomotor taxonomy which contained 13 classes: finger dexterity,manual dexterity, control precision, rate control (tracking), visual mem-ory, visual speed and accuracy, position memory, auditory discrimina-tion, auditory memory, clerical perception, depth perception, dividedattention, and kinesthetic memory.

A method was devised which would allow development of dataabout the amount of each of the perceptual/psychomotor classes involvedin the performance of the tasks of an AFSC, and the performance qual-ity variability produced of each of the classes when the tasks of an AFSCare performed. The data collection method was modeled after the jobinventory procedures developed and currently used by the Air Force.This procedure essentially involves a checklist of tasks which is com-pleted by a job incumbent relative to a set of attributes. In the presentcase, the attributes were the taxonomic classes and each list containeda sample of the tasks performed in each AFSC investigated.

A pretest of the procedure indicated that the method yielded datapossessing sufficient quality to meet the goal of the present work--toprovide a description of the perceptual/psychomotor requirements ofvarious AFSCs. The job incumbent raters (supervisors and subordi-nates) were able to understand the taxonomy, the scaling procedures,and the methods for responding. The results, based on data collectedduring the pretest phase, also suggested adequate test-retest reliability.Scaling procedure or rater variables (supervisor or subordinate) did notseem to affect the results. However, the pretest also indicated a num-ber of areas of potential problems. These largely concerned range re-striction and discriminating power. To cope with these problems duringthe major data collection effort, modifications were incorporated intothe methods.

The revised methods were applied to collect perceptual/ psycho-motor requirements information about 35 AFSCs at 10 Air Force baseswithin the SAC, TAC, and MAC major commands. The task list for

175

ADA093 981 APPLIED PSYCHOLOGICAL SERVICES INC WAYNE PA F/6 S/10PERCEPTUAL/PSYCHOMOTOR REQUIREMENTS BASIC TO PERFORMANCE IN 35 -- ETC(U)DEC 80 A I SIEGEL. P J FEDERNAN, E H WELSAND F33615-78-C-0032

UNCLASSIFIED AFMRL-TR-80-26 NL3EIIIEIIEIIIIIIIIIIIIIIIIIlfflfflfmIuIuIuIuIIuumIIuuIuIIIuIuu

IIIII 7

each specialty contained a sample of 60 tasks performed by incumbentsin that specialty. The resultant data were analyzed to yield the "high"and the "low" perceptual/psychomotor ability requirements for each ofthe AFSCs involved. Across all 35 AFSCs considered, the four mosthighly rated perceptual/psychomotor abilities for both amount and per-formance quality variability were: visual memory, visual speed and ac-curacy, finger dexterity, and manual dexterity. The four lowest ratedabilities for both rated aspects were: auditory memory, auditory dis-crimination, kinesthetic memory, and depth perception.

A high correlation (r = . 97) was found between the ratings of"amount" and of "performance quality variability. " This suggests thatonly one or the other of these factors need be included in future sur-veys of this type. As in the pretest, the correlation between the datayielded by supervisors and by subordinates was high. For the amountfactor the correlation between the ratings made by the two sets of judgeswas . 94; for the performance quality variability factor, the correlationwas . 90.

Indices of profile similarity indicated suppoi-t for a contentionthat the perceptual/psychomotor requirements for most AFSCs are atleast partially unique.

A reliability analysis indicated that about 24 raters are requiredto produce adequate stability within the techniques employed.

Sources of rater error- -leniency, contrast, and halo--were in-vestigated and found to be present but not to a large extent.

A factor analysis of the data indicated that the taxonomy can bedescribed by three factors--a visual, an auditory, and a manual factor.

Conclusions

Within the limits of the present data set, the following conclu-sions seem warranted:

1. The taxonomy, here developed and employed,can provide a basis for describing the percep-tual/psychomotor requirements of variousAFSCs.

2. The methods and procedures, here developedand employed, seem to produce information ofadequate quality and can be employed in futureinvestigations of the present type.

176

3. The various AFSCs are somewhat, but not high-ly, independent relative to the perceptual/psy-chomotor abilities required for effective perform-ance. No AFSC was free from perceptual/psy-chomotor requirements.

4. About 24 judges are required to produce adequatedata stability within the techniques employed.

5. By extrapolation, the results of the present workcan be employed as a basis for career classifica-tion and perceptual/psychomotor test development.

1i

177

REF ERHEN CES

Alluisi, E. A. Methodology in the use of synthetic tasks to assess com-plex performance. Human Factors, 1967, 9t, 375-384.

Bechtoldt, H. P. Selection. In S. S. Stevens (Ed. ), Handbook of ex-perimiental psychology. New York: Wiley, 1951.

Camrpbell, J. P. , Dunnette, H. D)., Lawler, E. E. , 111, & Weick, K.E. , Jr. Managerial behavior, performance, and effectiveness.New York: McGraw Hill, 1970.

Chambers, A. N. Development of a taxonomy of human perform-ance:A heuristic model for the developm-ent of classification systemns.JSAS Catalog of Selected Documents in P--sychology, 1973, 3, 24-25.

Christal, R~. E. The United States Air Force occupational researchpr-oject. Lackland AFB, Texas: Air Force Human Resources Lab-oratory, January, 1974. AFHRL-TR-73-75, AD 774 574.

Cliff, N. Scaling. Annual Review of Psychology, 19173, 24 473-506.

Cronbach, L. J., Gleser, G. C., Nanda, H., & lRajaratnarn, N. The

dependability of behavioral measurements. New York: Wiley, 1972.

Eisler, 11. On the problem of category scales in psychophysics. Scandi-navian Journal of Psychology, 1962, 3, 81-87.

Eister, H. Magnitude scales, category scales, and P'echneriari inte-gration. Psychological lleview, 1963, :t_% 243-253.

Eilerci, H. , & Montgomery, 11. On theoretical and realizable ideal con-ditions in psychophysics: Magnitude and category scales and theirrelation. Perception & Psychophysics, 1974, 16, 157 -168.

Ekman, G. P1sychophysik and 1'sychologische mfessu ret hoden. InMeili & llohraclier (Eds. ), Lehrbuch der Experimentellet' I syc ho-logic. Bern and Stuttgart:, Verlay Flaus Hluber, 19613.

Ekmian, GA., & Khnnapas, T. Scales of conservatism. Perceptual andMotor Skills, 1963, 16, 329-334.

Ekmari, 6. , & Sjoeberg, L. Scaling. Annual Reviewv Of PsyVchology, (195 j6 5144

Engen, T., & McBurney, D. H. Magnitude and category scales of the

pleasantness of odors. Journal of Experimental Psychology, 1964,68, 435-440.

English, H. B., & English, A. C. A comprehensive dictionary of psy-chological and psychoanalytical terms. New York: LongmansGreene, 1958.

Fitts, P. M. Factors in complex skill training. In R. Glaser (Ed.),Training research and education. Pittsburgh, Pa. : University ofPittsburgh Press, 1962.

Fleishman, E. A. Human abilities and the acquisition of skill. In E.A. Bilodeau (Ed.), Acquisition of skill. New York: AcademicPress, 1966.

Fleishman, E. A. Toward a taxonomy of human performance. Amer-ican Psychologist, 1975, 30, 1127-1149.

Galanter, E., & Messick, S. The relation between category and mag-nitude scales of loudness. Psychological Review, 1961, 68, 363-372.

Gibson, R. H., & Tomko, D. L. The relation between category andmagnitude estimates of tactile intensity. Perception& Psychophysics,

1972, 12, 135-138.

Gregson, R. A. M., Mitchell, M. J., Simmonds, M. B., & Wells, J.E. Relative olfactory intensity perception as mediated by ratio-range category scale responses. Perception & Psychophysics, 1969,6, 133-138.

Guilford, J. P. A study in psychodynamics. Psychometrika, 1939, 4,1-23.

Guilford, J. P. Psychometriu methods. New York: McGraw-Hill, 1954.

Guilford, J. P. The nature of human intelligence. New York: McGraw-Hill, 1967.

Harrow, A. J. A taxonomy of the psychomotor domain. New York:David McKay, 1972.

Horowitz, L. M., Inouye, D., & Siegelman, E. Y. On averagingjudges ratings to increase their correlation with an external cri-terion. Journal of Consulting and Clinical Psychology. 1979, 47,453-458.

180

Hunter, D. R. Development of an enlisted psychomotor / perceptualtest battery. AFHRL-TR-75-60 AD-A020 544. Lackland AFB,Texas: Air Force Human Resources Laboratory, November, 1975.(NTIS No. AD-A020 544).

Hunter, D. R., Maurelli, V. A., & Thompson, N. A. Validation of apsychomotor/ perceptual test battery AFHRL-TR-77-28. LacklandAFB, Texas: Air Force Human Resources Laboratory. AD-A044 525.

John, I. D. Stimulus discriminability in the magnitude estimation andcategory rating of loudness. Perception & Psychophysics, 1969, 7,229-234.

Kirk, R. E. Experimental design procedures for the behavioral sciences.Belmont, California: Wadsworth Publishing Co., 1968.

Lawlis, G. F., & Lu, E. Judgment of counseling process: reliability,agreement, and error. Psychological Bulletin, 1972, 78, 17-20.

Lopez, F. M. The threshold tracts analysis system. F. A. SystemBulletin, Port Washington, N. Y. : Lopez and Associates, SpecialReport No. 1., 1978.

McCormick, E. J., Jeanneret, P. R., & Mecham, R. C. The devel-opment and background of the position analysis questionnaire (PAQ).Lafayette, Indiana: Occupational Research Center, Purdue University,June 1969.

Miller, R. B. Task taxonomy: Science or technology. Ergonomics,1967, 10, 167-176.

Montgomery, H. Direct scaling: category scales, magnitude scalesand their relation. University of Goteb'rg, Sweden: GotebbrgPsychological Reports, 1975.

Montgomery, H. Magnitude scales, category scales, and the generalpsychophysical differential equation. Perception & Psychophysics,1977, 21, 217-226.

Morsh, J. E. Job analysis in the United States Air Force. PersonnelPsychology, 1964, 17, 7-18.

Nie, N. H., Hull, C. 11., Jenkins, J. G., Steinbrenner, K., & Bert,

D. 11. Statistical Package for the Social Sciences, New York: Mc-Graw-Hill, 1975.

181

Osgood, C. E., & Suci, G. J. A measure of relation determined byboth mean difference and profile formation. Psychological Bulletin,1952, 49, 251-262.

Pfeiffer, M. G. Mental work criterion development through scaling:A technique for determining the intellective structure of the collegeprofessor's job. Perceptual and Motor Skills, 1970, 30, 643-652.

Pfeiffer, M. G., & Siegel, A. 1. Comparison of category and magnitudescales of technical skills. Perceptual and Motor Skills, 1966, 22,235-248. (a)

Pfeiffer, M. G., & Siegel, A. 1. Personnel psychophysics. Percep-tual and Motor Skills, 1966, 22, 257-258. (b)

Pfeiffer, M. G., Siegel, A. 1., Taylor, S. E., & Shuler, L., Jr.Background data for the human performance in continuous operationsguidelines (Draft). Wayne, Pa.: Applied Psychological Services,September 1978.

Pradham, P. 0., & Hoffman, P. J. Effect of spacing and range ofstimuli on magnitude estimation judgments. Journal of Experi-imental Psychology, 1963, 66, 533-541.

Rabideau, G. F. Field measurement of human performance in man-machine systems. Human Factors, 1964, 6, 663-672.

Ragsdale, C. E. How children learn the motor types of activities, inN. B. Henry (Ed.), The forty-ninth yearbook of the national societyfor the study of education, Part 1: Learning and instruction. Chicago:University of Chicago Press, 1950.

Rarick, G. L., & Dobbins, D. A. A motor performance typology ofboys and girls in the age range 6 to 10 years. Journal of Motor Be-havior 1975, 7, 37-43.

Schneider, B., Parker, S., Valenti, M., Farrell, G., & Kanow, G.Response bias in category and magnitude estimation of differenceand similarity for loudness and pitch. Journal of ExperimentalPsychology: Human Perception and Performance, 1978, 4, 483-496.

Siegel, A. 1., & Federman, P. J. Qualification analysis for seventelephone company jobs. Wayne, Pa.: Applied Psychological Serv-ices, January 1976.

182

---------------------------- --

Siegel, A. I., & Musetti, L. L. Analysis of electromechanical switch-ing jobs. Wayne, Pa. : Applied Psychological Services, September1978.

Siegel, A. I., & Pfeiffer, M. G. Factor analysis of category and mag-nitude scales of a technical attribute. Journal of Applied Psychology,1966, 50, 341-347. (a)

Siegel, A. I., & Pfeiffer, M. G. Personnel psychophysics: Operationalcorrelates of electronic circuit complexity. Wayne, Pa. : AppliedPsychological Services, 1966. (b)

Simpson, E. J. The classification of educational objectives in thephychomotor domain. In The psychomotor domain: Contributionsof behavioral science to instructional technology. Washington, D. C.:Gryphon House, 1972. (ERIC Document Reproduction Service No.ED 084869.)

Sjoeberg, L. Unidimensional scaling of multidimensional facial ex-pressions. Journal of Experimental Psychology, 1968, 78, 429-435. (a)

Sjoeberg, L. The dimensionalitY paradox in comparative judgment: Aresolution. ScandinavianJournalof Psychology, 1968, 9, 97-108. (b)

Stevens, S. S. On the psychophysical law. Psychological Review, 1957,64, 153-181.

Stevens, S. S. The psychophysics of sensory function. in W. A. Rosen-blith (Ed.), Sensory communication. New York: Wiley, 1961.

Stevens, S. S. The surprising simplicity of sensory metrics. Amer-ican Journal of Psychology, 1962, 17, 29-39.

Stevens, S. S., & Galanter, E. Ratio scales and category scales for adozen perceptual continua. Journal of Experimental Psychology,1957, 54, 377-411.

Torgerson, W. S. Theory and methods of scaling. New York: Wiley,1958.

Whitlock, G. I. Application of the psychophysical law to performanceevaluation. Journal of Applied Psychology, 1963, 47, 15-23.

183

Winer, J. B. Statistical principles in experimental design. New York:McGraw-Hill, 1962.

Yoder, D., Heneman, H. G., Jr., Turnbull, J. G., & Stone, C. H.Handbook of personnel management and labor relations. Nev York:McGraw-Hill, 1958.

184

APPENDI( A

Pretest Forms

185

a

SECTION I

AMOUNT OF PERCEPTUAL/ PSYCHOMOTOR ABILITY

Part 1

Your task is to judge how much of each of a set of perceptual/psychomotor abilities is involved in a number of the duties in your ca-reer field. To provide this information you should:

1. Make sure you understand the definition of each abilitywhich is defined in the "Perceptual/ Psychomotor AbilityDefinitions" list.

2. Read the first duty in the duty list (last page of this 'Part").Refer to the first ability definition, finger dexterity (FD),and decide how much is involved in performing the firstduty. Enter your estimate in the first column. Use theamount scale below in making your judgment.

Amount

N =duty not performed in your squadronNH = performed, but ability is not required

in duty performance1 =very little (0% - 20%)2 =some (21% - 40%)3 =rnoderate (41% - 60%)

4 =considerable (61% - 80%)

5 =very much (81% - 100%)

3. Repeat the same step for the second duty, and so on untilevery line in the finger dexterity column is filled in.

4. Evaluate all the duties for the second ability, manual dex-terity, in the same manner. Enter your judgment for thisability in the second column.

5. Continue in this manner until every duty has been evaluatedfor all the abilities.

6. Refer to the ability definitions as often as necessary when

making your judgments.

Instructions for Parts 1 and 2 in Section I of the pretestdata collection forms.

187

-mow

7. An example of how to complete Part 1 is presentedbelow. For the first ability, finger dexterity, theperson making the judgment thought that "conductinginventories" requires "very little" finger dexterity.He considered finger dexterity to represent about10% of the duty. Accordingly, he entered a "1" onthe first line of the first column. For the secondduty, "maintaining files," the rater thought thatfinger dexterity represents 35% of "maintaining files."Therefore, he wrote a "2" alongside the second dutyin the first column.

Perceptual/Psychomotor Ability

11 21 3 41 51 V6 7 1 8 91 101 11 12 13Duty FD MDCPRC VMVS PM AD AM CPDP DA KMI

1. Conductinginventories i 2. N9 3 3 3 1NR NP, NI R W

2. Maintainingfiles 2 % \YR 3 I NRN NN R

8. Haters sometimes make errors of judgment when usingforms such as this. To avoid these errors:

a. use the high and the low numbers (1 and 5) ofthe scale whenever Appropriate

b. avoid excessive use of the middle number (3)of the scale

c. avoid assignment of the same value to allabilities in a duty

9. Enter "N" alongside any duty that is not performed in yoursquadron.

10. Enter "NH" in the box for any ability that is not required inthe performance of the duty.

188

Part 2

In Part 2, your task is to judge again how much each of the set ofperceptual/ psychomnotor abilities is involved in a number of the duties inyour career field. To provide this information you should:

1. Make sure you understand the definition of each abilitywhich is defined in the "Perceptual/ Psychomotor AbilityDefinitions" list.

2. Read the first duty in the duty list (last page of this"Part"). Refer to the first ability definition, fingerdexterity (FD), and decide how much is involved inperforming the first duty. Enter your estimate inthe first column. Use the amount scale below inmaking your judgment. You may use any value fromzero to 100.

Amount

I i I I II I I I I I I I L I I0 10 20 30 40 so 60 70 80 90 100Very little Some Moderate Considerable Very much


4. Evaluate all the duties for the second ability, manualdexterity, in the same manner. Enter your judgmentfor this ability in the second column.

5. Continue in this manner until every duty has been eval-uated for all the abilities.

6. Refer to the ability definitions as often as necessarywhen making your judgments.

7. An example of how to complete Part 2 is presented below.For the first ability, finger dexterity, the person makingthe judgment thought that "conducting inventories" requires"tvery little" finger dexterity. Accordingly, he entereda "10" on the first line of the first column. For the secondduty, "maintaining files," the rater thought that "some"Iamount of finger dexterity is involved. Therefore, he en-tered a "25" alongside the second duty in the first column.

189


1 21 C3 4 1 5 1 6 71 81 9 101 111 12 131Duty FD MD P R VM SA M AD AM CP DP DA 0

1. Conductinginventories 10 35 140 NR (pO 6-2 NR pR 2 KR RAI

2. Maintainingfiles 1, 00 55 X1RWN,&F V

8. Raters sometimes make errors of judgment when usingforms such as these. To avoid these errors:

a. use the high (80 to 100) and the low numbers (0 to20) of the scale whenever appropriate

b. avoid excessive use of the middle numbers (45 to 55)of the scale

c. avoid assignment of the same value to all abilitiesin a duty

9. Enter "N" alongside any duty that is not performed in yoursquadron.

10. Enter "NR" in the box for any ability that is not requiredin the performance of the duty.

190

SECTION 11

PERFORMANCE QUALITY VARIABILITY AS A FUNCTION OF

PERCEPTUAL/ PSYCHOMOTOR ABILITY

Part 1

Your task in Section 11 is to judge how performance differenceson a particular duty vary as a function of a specific perceptual/psycho-motor ability. In completing the form, you should ask yourself the ques-tion: For a great number of people, how large are the differences inthe performance of the duty, as a function of this specific ability? Yourresponses will provide information about the quality of performance,over large numbers of people. To provide this information you should:

1. Make sure you understand the definition of each abilitywhich is defined in the "Perceptual/Psychomotor AbilityDefinitions" list.

2. Read the first duty in the duty list. Refer to the firstability definition, finger dexterity (FD), and decide onthe extent of the performance difference which existsbetween people who perform the duty as a function offinger dexterity. Enter your estimate in the first col-umn. Use the performance quality variability scale be- Ilow in making your judgment.


N = duty not performed in your squadronNR - performed, but ability is not required in duty performance

1 - quality of performance of this duty varies little, because ofdifferences among current personnel in this ability

2 - quality of performance of this duty varies some because ofdifferences among current personnel in this ability

3 - quality of performance of this duty varies moderately becauseof differences among current personnel in this ability

4 - quality of performance of this duty varies considerably becaiseof differences among current personnel in this ability

5 = quality of performance of this duty varies to a very gieat extentbecause of differences among current personnel in this ability

Instructions for Parts 1 and 2 in Section li of the protestdata collection forms.

191

wI

.. . .. , ..... . . .. . .. .j


4, Evaluate all the duties for the second ability, manual dex-terity in the same manner. Enter, your judgment for thisability in the second column.

5. Continue in this manner until every duty has been eval-uated for all the abilities.

6. Refer to the ability definitions as often as necessary whenmaking your judgments.

7. An example of how to completePart 1 is presented below.For the first ability, finger dexterity, the person makingthe judgment thought that quality of performance of theduty, "conducting inventories, " varies "little" among cur-rent personnel because of differences in finger dexterity.

Accordingly, he entered a "1" on the first line of the firstcolumn. For the second duty, "maintaining files," the raterthought that finger dexterity differences among currentpersonnel have "some" effect on the quality of performance.Therefore, he wrote a "2" alongside the second duty in thefirst column.


1 2 3 1 4 51 6 1 7 81 9A0M12 1Duty FD MD CP RC VM VSA PM AD AM CP DPl DA KMK

1. Conductinginventories t R33 3 _ N PI

2. Maintaining P4rR ig 10,_ Ifiles - 1 . it --1R 1. 1 _M NR RR41

8. Haters sometimes make errors of judgment when using formssuch as these. To avoid making these errors:

a. use the high and the low numbers (1 and 5) of thescale whenever appropriate

b. avoid excessive use of the middle number- (3) ofthe scale

c. avoid assignment of the same value to all abilitiesin a duty

192

9. Enter "N" alongside any duty that is not performed inyour squadron.

10. Enter "NI" in the box for any ability that is not requiredin the performance of the duty.

Part 2

In Part 2, your task is to judge again how performance differeniceson a particular duty vary as a function of a specific perceptual/psycho-motor ability. In completing the form, you should ask yourself the quem-tion: For a great number of people, how large are the differences in theperformance of the duty, as a function of this specific ability? Your te-sponses will provide information about the quality of performance, overlarge numbers of people. To provide this information you should:

1. Make sure you understand the definition of each abilitywhich is defined in the "Perceptual/ Psychomotor Abil-

ity Definitions" list.

2. Read the first duty in the duty list. Refer to the firstability definition, finger dexterity (FD), and decide onthe extent of the perfornance difference which exists

between people who perform the duty as a function offinger dexterity. Entel your estinlate in the first col-

umn. Use the. performance quality variability scalebelow in making your judgment. You mnay use any

value from .zero to 100.


l iii III I I I I I 1 I I 1, Iii I

0 10 20 30 40 50 60 70 80 90 100

Little Some Moderate Considerable Great extent

3. Repeat the samie step for the secotid duty, and so oil until

evel,y line in tilt finger dexterity colullill is filled in.

4. Evaluate all the duties for the second ability, manual d x-terity, in the salle imalnner. Enr e youri judgment for this

ability in the second colunn.

193

5. Continue in this manner until every duty has been evaluatedfor all the abilities.

6. Refer to the ability definitions as often as necessary whenmaking your judgments.

7. An example of how to complete Part 2 is presented below.For the first ability, finger dexterity, the person thoughtthat quality of performance of the duty, "conducting inven-tories," varies a "little" amount among current personnelbecause of differences in finger dexterity. Accordingly,he entered a "10" on the first line of the first column. Forthe second duty, "maintaining files," the rater thoughtthat finger dexterity differences among current personnelhave "some" effect on the quality of performance. There-fore, he wrote "25" alongside the second duty in the firstcolumn.


1 1 2 3 4 5 6 7 8 9 1 10 11112 13Duty FD MD) CP JRC VM VSAJ PM AD AM CP DP IDA KM

1. Conductinginventories % >5 O N 'C 52 L NR NR 23 IW K

2. Maintaining 5 11 MR 5 50 5 W t4R 40 O NRfiles

8. Raters sometimes make errors of judgment when using formssuch as these. To avoid making these errors:

a. use the high (80 to 100) and the low (0 to 20) endsof the scale whenever appropriate

b. avoid excessive use of the central area of thescale (45 to 55)

c. avoid assignment of the same value to all abil-ities in a duty

9. Enter "N" alongside those duties that are not pteifor'med inyour squadron.

10. Enter "NIC' in the box for any ability that is not requiedin the peilormam'e of the dut.y.

194

The respot,.se sheets were arranged with the 13 perceptual/psychlomotor abilities in vertical columns, with legends across thetop of tile page. The tasks A'ere identified along the left side ofthe page. A sample response sheet is shown in Exhibit A-1. In orderto balance the effects of fatigue and practice, the sequence of tasksin each of the two parts in each survey form section, were reversed.Accordingly, tasks ordered from 1 to 20 in Part 1 of the two sectionswere ordered from the twentieth task first to the first task list inthe second part of the survey form.

195

Cdd

I

i 1;i I

19

I 0

APPENDIX B

Interview F ovrm Used In Pretest

19J7

E7I

PERCEPTUAL/ PSYCHOMOTOR

REQUIREMENTS INTERVIEW

N ame Date

R ank________ Squadron________ Location

Career Field________________

I HAVE A SHORT INTERVIEW WHICH WILL, IN PART, BE BASED

ON THE FORM YOU COMPLETED. YOUR ANSWERS WILL BE HELD

CONFIDENTIAL. THE DATA WILL BE TREATED STATISTICALLY AND

YOUR NAME WILL NOT BE ASSOCIATED WITH ITr IN ANY WAY. THANK

YOU FOR YOUR COOPERATION.

1 99

1. (show card) How long have you been in (career field)

(1) Less than 6 months(2) Between 6 months and I year(3) 1 to 3 years(4) 3 to 5 years(5) More than 5 years

2. (show card) How long have you been stationed at ?

(1) Less than 6 months(2) Between 6 months and 1 year(3) 1 to 3 years

(4) 3 to 5 years(5) More than 5 years

3. (show card) How long have you been in the Air Force?

(1) Less than 6 months(2) Between 6 months and I year

(3) 1 to 3 years(4) 4 to 5 years(5) More than 5 years

4. (show card) What other per-ceptual/psychomotor abilities that werenot included in the list, are required in performing thethe duties listed in the questionnaire?

5. (show card)

5. (show card) Are there any perceptual/psychomotor abilities that maybe required at some locations or duty stations that wouldrot be required at others? Which and how so?

200

6. (show card) Which perceptual/psychomotor abilities were not

meaningful to you? 1low so?

7. (show card) Which two statements best describes your opinion of

the categorical questionnaire (identify)? 'he magni-tude questionnaire (identify)?

Categorical Magnitude

A good approachUnclearDifficult to seeProvides inadequate

information

Easy to use

Poor approachHard to understandProvides a lot of in-

formation

Easy to interpret

Clear

8. (show card) In general, how difficult was it for you to completethe categorical questionnaire? 'The magnitude

questionnaire?

Categorical Magnitude

Verv difficultModeratelV difficultNeither difficult or

easyModerately easy

Very easy

201

9. Do you have any suggestions that would be helpful to us for im-proving the form that you completed so that it will be more use-able ?

10. Please provide a judgment indicating the amount that each perceptual/psychomotor ability is involved in each of the following duties. (Pro-vide definitions. Enter responses on the answer sheets for- the inter-viewee. Within each career field half the interviewees use the cate-gorical scale and half use the magnitude estimation scale).

11. Please provide a judgment indicating how large the differences inperformance varies as a function of the specific ability. (Providedefinitions. Enter responses on the answer, sheets for the inter-viewee. Within each career field. those interviewees who receivedthe c-ategorical scale in item 10 should receive the same scale forthis exercise. Similarly, those who used the magnitude estimationscale in item 10 should use it again for this item).

202

AllttPI)DX C

Correlationls Aniong Abititics 1 01' Protest

20:1

I; s . 4 0 t. 0 0 o v P.~c; ,

Lr '~TC '0 % -. ~evJ- '0. 0< T ~ - iNJ Uf - r W pn' Fn C) Pd 0 ,

re) en o: C A 'T 0, r1n 00 '0 (\i

oL a ~ .0 . 0 C.

0 01 C) C 0u 0.) Cr'j03 '0S La 5 5 S S m

00 ri- 00 ~o r 00001

ev ~ ~ ~ ~ - 0.- 7 ) tCi nP1C

0)0;881 8848ccca

o-Cr ' - oCLn wv -j C) pn 'nr

0

oc

$4 000 0 '- Q00 0000 ~nC o-

o 4

-4 0 UC , n a C.C

0 on P- Fn D osUNo

-4

ca9.; 4C ; 8 C

4-1 ) *3 C) 0 C> 0 C)

0n V'4 r- co rv (U w(U L - (-.J

L- 000 00 r 000

Oc P ' 1 It (U 00 9Si. a a 8 gc gcC31 0C

2Gn 5 0 o 0 - "f

k^ 0-N.- ) a - '-0 C0C)0 NO C

W:l n. C! 0-t '7 n:'3 U n '

00000000 tv000 0 a -

(D 10 N ru - pn ~ 0% 'r N C'0 a. >

- R t 0 r e u g , ' 0 a0 - i ' "000D COO 000 C)C -0;C' Da

0nW 000 SmL % r- r--0 r Dr '

U'P L% u' (. -4c N b a r.- 'o unin

*d C) Ci C

0~ 0: C; C, ,-,

0 ItL 0 Ol O:Ln N s0L r,o ' * * * . * *

04c ~ c ('~ ) oC )C )cl

0 0cc0~-0 000D

Cd

04J 0 '~T" fn r- C:) W. ) .-O'' 0'n t^ ru

C-i C) M *n * ) C) n C) * 0 (n C *0

0

0

-:1 'r 'C ' '00' ~ N

0. : fnJ pl 1 '- 0000 or 000 0

5 C; ; C; *) ') Do of 0* C, 0'

, J 000000 00000 )r-n

206' .?

02N O'N j 0r c Dc C4 C C C', 0'M LA Ln D~

0O cc. nc n 0-l M -Zr C '4C eo-e

-4 C C OC *t

C) C8U cc

.4

Z ~~ C)0-

M~C '

a) 44 t: O -4c

o) 0

C C r C %C - , '-i %D . C,0

-~ CjUi - C'4 CND C) 0

C- C CC4 O CCm Cc 1 C r- a cI

-40 - - r(1C

T Ir C) enI 1 - - - ,

41 207

04 %0 -4 -n m~ VI 0, Ol 0) '0 Pn* r- 0

' -0 '00 -4 rJ -0 Lnu r -0 FI UN

0 0 0 * al * a . * S a0 4 C 0 00 C0 aCD0

,ac~ o v o 'Ooc -o 0 .

C.) ; ; . C* * C)" o 0* CD 5-* n, *D

'0 to 0000)0 CD: Lr 0~ )C o

000p0o oc0CDc J

.- P-w ~ - - C' 3 ON0- M -n 0

.- 4

a)0) C; C C C )C )

'T N 0a'oCI'rO -CO mj 000000.0; 00 ;

u 0

bc4 4o.), C c ,- L o0

E- 0

.0

I ~ ~ C (\I -O j V 0 %0 0.- N- C'- - I

0

0 C o C; n- a C 0 oCo o0

0 L

414(a C) C; D - C)CD C)C:) 0 0 C

40) 0 0 pn 5

. .. . Z -0)0Do 0 0 0

20

m~ L% o r. - e: o r-- r- w 0 r- C

-. 7~~~ .7 N u

0 C C QC)00 000~ 0- C)

o) C ) o- D 0 C 0 Q'a0 0 (

U)1 4

0 PL

-0 N w) a, C j

V0

w 4

a))

o CC' C) k) U *C L . . *

0 V C C$3$*r C) C)

-0 0 C C- CO~ - C) 00 C-)C-

41 L) \j t

0

u-0 C) *1 C- ?- *j C: *

0 $4P U \ 0 4 0 'r j r ' -J P)

C: -000 c0 000 CI :

C: C) VIN VN~ rNi I - ) 3- C) N~

-0 CC 0 0 0) 0 C C) 0 C) C)C

Li -

-Apo, r ~ . -.

~~.. .i .it~ . .O' .C . .N '0.

a, mi o~ r- 3 -a r- 3 N0

C\j '0- N Ni) N .- -3 '0 C* N

.0 vN 7 - .(,C oL

L) Cj D CD0 D ~0 0

r\ N. C3 oa -.t u'o C) cy -j fNi a

LI) .- 4 f, o cjr- 3ccJo L) r-c - nd

0 Q4

'.00 ; n c

A 4 W0 -

'A~~4~'DC-0 4-C

r4-

0- 9 X.r' N.C1;/1~'

0 Z7; 0C C) C7 Ci) , OC I-) C' C

0o- w r,' r- L) D N IN .

-4w

0 Tr NCD)N '0 C30O '

w C- 10 r- L- 0N-0r- z L ,

1- 0 0-)r C) CD U CD C3 Cl C3 C) C)

C) r- C-1 q )o cr% i io

j - - C)C L-3 C C CjC f l I c-) CJ C)

210

y0, CL cr, 0.) o 4,. c,; .- j o CcCC r~j Nj ('4 ('J~ -4

o 0 * - c * ) c C .

.- & ir ~,r.,~v*' CD- WN CD C

0 r 0, C. , co *

r4 .~. 1% 1- C)C) C .

p SD * * cy *) CD u 0

Q) I

C) .. C; 1 C1:- Y0 r- C- C \D '

(14~ c c' _- :c ( . ~L'.

4Jci

4..

r. r-. : !:E C l C - C CCZ O

., c' c -4~-'~ ~ - N'Cnj r,

0o

CO\rrl, I

C-. L7 CC3.

0 -,1 )

*-4 4

to s )

0o

Co C70 -7 'C C C1L :

r L

41

L -4

211-C~r

U o 0' - ('- o '0 O C) C-0c :tO - C) - - C) -r 'r ruj 0 '-

c)J C) C) C- 0- 0 C) 0A -:t 0 -0

0 I

~0oo

0 fl r ' - Q C

0D I30 M0C))C

I C,.7 I . !

4j E

co0U r0 1 0t ,CD i -.7 C'(\ 0-0 o CD r( LA

U-4 r% 4 ) I

V * . 0 . . . * .

ol 0 s0)00c 00i

0

co

4-j

1-4oQ)w . 4'G:::~(40.rJ4

4-1

'0 W% .- -. Mi. C o \j -o0 C M10 V. -S~' '0 ': M -n ~n(jW

n -0 00 (I0 r r)C% 0 000~ L

C) n CV- ) ' C)~ C) C- '0' CD C \I C)

212

), ~o LA LA i*' -, V -I f0 C71 a

0Cfl 000 000 0,

,. n Cu I j 0 '0 r Uo-N e-J 0 0

f". ~ F- 'o IC~ A. -c~

0 ~ L a 0 C C30

r. A0~ 'J. J crn 0 r\j 0~l C~

Mo 3 . . .t . . . .*

q Cl0000C: C OOL-C CD - CD00n

CO0000~O

0w

-H 0u4)

I) 41' D S 1

0

01o m x C; -C 0 0

.-4 cd * f

C13

- 4jcc, C c- 0 JciC C c4- 01 '

C)cC . 0 ci Jc. c- )4,c c,

1.40

C) )

0: C, i'- -7 ..T Ci)l ' r~- ci" -.1*J

C,- C'. c 0 C) C, C-)C CC)

C- ac cru r"- t'- w - 0

.- CD 0 C, C-) CD cJCL)-1 c 0 C)

2 13

0 4 f" n~0 1'0 u"' 1%. 0' v.it

cD~LI' is. o 0 o . Uo L') U' ' . 0 0

O - '0 .- -N ~-~d . n'0

0 - )0 D)DZ

0 0-0D0 0 0 D0 ~0 Cco

-4

0n ):00C) C) C7 C)

U. . . *

C 'D

0-0'0 O' ^- CD C) rl- -4

LW C 'r C - CO .

4-)0 C- CD 0 -CD 0 000 rj )r-

.' <r w c'".C 0jfr--.I'L(''a0 -0 V,o ;> C C C :,

s.4 C0 00C CCJ0

o0.

0)0O 0 ooo- ID0 cc nvi17 oo r

EA00 t)0

-I\ N C :) r P-L c.0 0

P4 . . . . . . .

0D CD- C) 00 0 0(,( 0 CD 0

O' 0 C) '. (\j f 0'c (Ln 0 (,\'

411

0, 0CD 0000000 D )

,O (I C' W) 0. N 0 ' (\.j 0 M

U' lf L^ fl- Ln r- PQ 0 '0 010 1

C) C 00 00;0 1- 0

0r. '.t N U- - N . r\J pn 0) 1 0 10

I4

It - L 17o. 0 r-rC)Cl IC K

<U

q 0C)00.-C) m CD)C 0C11Cw

0 4 C.. N: 'O - Cu' 1' J\U9

o P

4- 1

4.

4-j4

.) .L . L . . x . ' . s . o . . '

CO Cl C C) C) LD CI CD C-- C CJ

0) >

Q) ~ C r- Z , r, N D cc N \ N . Lr L/-4

0C C C C CD . C) C-- C) 0-

21

0:0 0 0 0 0 r :P CK00 0 0.

-: C. 0'0 CD 0 0 0~ 0'0 0 (D C

0D 000 00 C; 00 .- 0

41T W)

0 UW'o C) C o s

A4 .D . tn I' of * o 0

4 0 0 0 0 0 0 0 e-DCD DC0

4 .- 'i\I' . 0 X0'

'0 U

ito -n'r wt 0 ..i 0- N

Li 0

r-I 4

J-H

0 -

E 4 *4 >

0L >. > 1

U) r- " VF-OIL -

-14 1 J 7OC000 o zo ,41

0 4

41

it' f- r'i r i Pn - C . ~4 C0 C C C C 4D CD C,

141

0 CC C)O CD 0 - 0

o C; .; . .; . C C'

cfn CDC) ZC c-t 7.0 CD)

.,4

-.V cJCco9cc-D-c

-H J

.4 .~ . . .~ .A . ..

ci c GC c3 L-,. - , -c.jc:':0

o 0 . ...

-- 4 WI

o 44 u ,

-- i r- C~c -9- C C-) CC

01.

0. C~ C .- ~C, - , C

0 0

4 .C) C. C. C L) C-) 9- U.. CD C) C0

C (...r .1' r-. -,j tN fr -J C. I.

C . . ( .- . V. . . . .. . .N

C- C-) C

217

OT

- 0, , . 0 0, w .V ~ C3

CI ~r0 C.O C) 0 C)' C) CDC i0 Ci

.- ~ ~ . .ar - , v

00

4

cu

0)2

Co0 . . ..

.4)

ca * *4.

-4 >

U i..q 4-i C) 4 ~ C1' - 4N i)

CiiC) 0; ) )D i- , C) LC)C0

.0 r-I

0 0

4- 4)* a ;

Ip. C-C) 0.- Y l QGC L"JC I, r,

i-i ~ ~ ~ C CC.c iC)OLC')O )C

2 18C'j?- x.- ' - - ~

C)C)t C C) M C) ( 0.

C) C) C)0 C C)0C C-) C. CD

4 U' 0 N)N U0 L' N- C C -( i Ln)

(:JC) C L) C C)i -D C) Ci C)~ CD,Q0 )

*w*4 -"C 4N- N- N- U) CD -

0(1, .; . . . .

00'

-4 41

) I r- P xL. . '.c. . . C . .P C r . :

$4 (D0 C::C-' C) 0C CD 0 C)C, 0.0o 00t

E-4' 444C C~N -Cr4 o. 1: - C l1

0 0.

044 rrf-p 0 ',c

(a- 4 - 7: 7-OC-- I C' ) C

0404 DC3 .l.Z- r l QC C)Q )C

41 CD C C)C ) )C C) C D 0- C

C.4 C)~ ) L/ tC\ . - i

* . . C P" "CC4t I '

iL4 c0' c) C C))Ci c, l

219

0 ' 0 0 V% a.- -j Z3 a, '0 CDIt . IT. 17 r'j wl W) l co 'IP CD

00 0V 0n C) ) cn z 0 'r.

C;' (a)' 0 ) 0' 1 0 0 Cr I D C) C)

A- sn -0 -N 'o 'd x\ .ru _0 cr. c-

0J

0

o ' cil -j C,- -- I a C

41. . . . . . .

~J W .' 0 " 1i - 'T cj C . rsC aj .D-2

c r4

* * * * * * * * 0 - *0 C, C3 C CC a- D CJ C-)U

0.

0 1N .rj C-1 N V P,-- C

tW W i>

U L5-4 4J

000 Cd-$)~)CCCcrC

E ? -,C0

i0

f) (- C' C1 0 r- C)rl C-1 1 C)

C)(t'. C- ~ ) C'N (Xi -0\ I 0Cl . 0. .- .4 r-- C. C '' '

C-) C ) 0 - C) C) C) (D C) n C i C-1 )

220

APPFNI)IX 1)

Frequency Distributions of Amount Ratings for

13 Perceptual! Psychomotor Abilities

221

rill!

5. ao-5.99 1

- I I - -

000-1.99 35 9 0 7 53 0 197 0-

Z9 004 9 36 36 1 17 19 a1 13 0 125 7

1 .l - .9 9 -o 4 0 1 0 -

09-.99 0-0 3 ~ - - --

1

3.-3.9 0 0 6 1 10 19 9 3 4 3 6 3 212

1.3.199 10 10 14 13 '0 39 09 3 3 0 30 17 Z0

.00-0.99 13 19 1 - 33 06 s

47

5.00-5.99 0 03 1 13 9 6 7 3 11 16 1 -

1.0-3. 99 to 98 2 A 0 1 9 a 3 A 10 z 19

24o- .9 50 1 7 1 1 1 16 3 0 13 1 1 a 0 3 0 3

l.0-.9 is 30 k t9 10 1o L 9 0 03 35 0

-- 99 .1.

7 6 3 3 16 3

3:00-3.99 19 1 7 30 04 -3

10-1.9 L 1O 39

. 9 1 0 7 8 1 36 3 1

9~ 13 . I -0I

3.0-99 20 20 14 - 9 1 7- 3

0 .00-.99~ _5 164 7 11 3 19 99 0 10 0 17t

3,0199 6 30 19 t6 9 9 0 1 0 7 1 6

330-3.99 1 *t 9 .9 ~ ~ 5 1 9

30-3.99 5 0 1 21 1 9 0

.30-.9

)0 i.99I

3 IIs

32 7

6 b1301.q

5 32 09 I 13 9 - 10

393.90 5 4 01 I

>301.9 7 5 1 9 I l .1 002239

3 o0.. 99 - - -3-

.30-1..99 - 3 5 1 1

.30- .99 .3 22 . 3 .A 33 10 1 5 4

2.30-0.19 1 . 3 5 3 2 - o 3

Aa. S. t9..

3.0 .99 1 I -o - 3 - -L.. o30. L L I - .- - 11 15. - 2

330-3L.99 9 A I 3 -: - 0 - - - .1.30-).09 :3 13 29 2 5 25 2 3 2

o30-.9 -1.0-3.44 11 - 11 -1 3 2 . - 3)30-.9 M 0 3 3 2 L 2 o

6.30-0.99 3 3 23 - - - . 2 2 5 - - 0

30-o0 1 -1 1 Is - - - -

3: 30-3. 99 10 - - - - --

..3t2. 9 1. -1 - - 25 9 - - - 23 - - -

2. 0 2 4 23 3 24 2 24 20 11 -3 - - -2.3-219 36 23 .3 30 - 22 - . O 5 2

JO-0 9 - o - 36 28 - 9 0 - - 8 O .

3- .99 - -4 7 93.3-3. 9 2 - - 6 58 34 -6 51 - 2

2.02.9 93 4 .3 2 0 3 2 16 1 12 8 32 30o-,.99 39 20 .0 3 -0 2 3 3 4 52 38 332330-.49 29 : 2 - 4 2 1 4 *

!AQrVa AA, 0Oforatto.

3 303.99 3 - - . - - - - - . .

300-3.99 2 8 9 2 9 4 - - 29 4 - - 0

2 0-.4 2 0 2 9 2 9 2 - : 0o0-.9 - 9 9 . 4 2 3 2 3 3 3

.30-,3 - - 3 8 2 3 9 3 .

224i

wt 4D cr ac IN8 4A6 F49 AD M69 _P DA . SUM30

'18.Lbr r~r.-.

..0-.99 - 1 15 6 38 - - 113.3W-3. )9 .1 - 3 z1 ,4 1 - - I I S - 0

k:.30- L.9 L 9 L 1' -9 k l.. .

1.30-3.49 13 -2 5 z S3 53 6

5.10-3.39 1 - - 3 . - -2

1.01.99 i5 : I 16 20 3 - 1. -- -

.10- . 09 is i5 9 6 1s 32 1 5 13 6 , o;~-.9 .3 50 16 - .8 6 - 55 52 32,

.30-..)9 4 -d - - 13 -1 1

- - -- - - -I - T,21.~3-3,a 3 - 1 13 IS I 1 - . - 9

:)0-,).99 .3 33 1 3 - 53 39 5 1: 592

3dJ.99 A i9 i3- 1 3 9 5 II 0 6

1.30-1.49 in 11 1 15 16 28 2? 3 8 32 1903.0-.9 38 . . I O .0 2 2 3 Z3 131

030- 9

31~ .9 50 52 33 it 3; 38 1

i.0-.9 32 32 19 17 . 1 il 13 i 1 - .16

130-0.49 1, 4. 9 : 1 6 1 2 2 532

t1.?rcc 03.0-99- - .'5S-

10.9 13 3a 16 3 3 ~ 3 . ' 8 1

3.301.99. .- .6 1? 9 2253

vsS;.1. FO 30 C? -- C v33 &A PH AD AM6 c? IV DA .14 SLO

5.10-.99 - i - - -L at- -

a.0-53.99 1 17 9116 1 0

I1.00- 1.9 i0 13 16 t3 1 1 3 11 30 13 . 1 .430

3. J0-).99 - - 11I 1 1 31 40 lb 33 .5 bP 34

5.00-5.99 6

..00-1.9 6 IS 2 - 6 9 1

.10-3.9 7 1 0 - 1 3 1 3 -- - 16

2.0-.9 13 44 1 1 1 10 16 1 1 6 17 191 .00- L.99 t- 'I to21 I 9 30 a 4 3 6 4 0 3 43.30-0.99 - 13 37 L 1 1 4 32 19 1.1 3 177

Aircraft Electricfl I .... .Spc.10.c

5.30-5.99 - - I - - - - 1

.0-.9 3 11 3 1 19 3 32 - . - - -9

3.30o-3.9 23 6 16 2 9 6 23 1 - 3 - - 5 1.3-.9 19 19 23 9 11 6 9 6 3 1 13 166~l

1.30.;.99 17 '1 37 9 46 4 1 19 3 t53 113

3.a0-0.99 L 9 S! - 7 1 -3 . .

5.00-5.99 5 t; 0 3 - - - - - - 3

1.029 7 3 t' 16 1 31 16 3 1 16

I.003 1- 3 3 3 3 . -

5.0-.9 13 p3 3 1 1 t9 - - 7

1300-31.99 3 1 6 1 3 2

90.0 d .

9.0 1 0 1 t 10 4 1 4 3 15 s0 a 1 p 166

1.0-L.09 16 - 3 6 6 e 1 1 1 7 3 1 11O.77-0.99 - - - - - - .- : 1 10

51eaI0AI to.. Pr.Icpo

.10- P9 6 56 162- . . . .

3.93.9 3 31 26 9 30 a 3

3.03.9 35 3 . 11 2 14 9 - b Ise1.

".00-.9 3 9 13 1 4 19 2 139 3 1 3 13 . 3 16

.30-~3.19 .4 0 6 - - - 3 9 1 9 3 3

AilIrcat l I V rn. MCh-L,

IG, 1 .- 13 .Cp 9 . I3

3.0799 .q 39 .3 1 .3 . 9

0. 0 . 9u I At p p p e t c

.30- .9.9 - 3 . -

9.333926 P3 9 . .226.

AIPPENDIX E~

requency Distributions of

Perifor mance Quality Variability Ratings for

13 Perceptual/Psychomotor Abilities in 35 \F'SCs

227

I -

R3 7 1 C 3M AD AN 6 .0 CF )A F4 S0UN7

777CR924ICAL

J.O-3.09 5 1. 3 2 2 - - - 3 - 2-. 30-39 3 1 6 2 26 3 - 3 3 3 151.30-L.99 34 33 .3 2 39 3 3 11 20 32 U2 .6 11 35

-0--3Q- - )9. 40 03 42 1 69 243

!.a0-3.49 15 16 2 9 27 2 - 1 - - 032.00-2.89 29 26 38 16 9 23 22 1 - 6 331 32 8 35

2.3-1.9 3 1 2 23 .3 9 12 .0 3 9 13 2 6 2 330-39 - 19 21 3 3 2 13 67

Air?...a.r aO .3.0CarO 3.c0013.*

5.00-5.99 - - - - - - - - - 32 2 3 a 3 - - - . - - - 2

.3 39 9 13 L7 19 298 1 6 9 1 3 4 - i672.30-2.49 .1 32 17 26 23 22 37 0 25 13 33 30 32 36L .30- 99 5 16 23 14 3 3 i7 31 27 2 N 26 6 28 2421 .30-0.99 - - - I - - 3 1 - - - 3

3.30-3.99 -1 .6 -3 - I- 31.00- L99 14 13 34 3L 33 13 17 33 43 40 47 39 36 .153. 00- 3.99 5 L - 7 3 1 0 4 26 3

3.00-3.99 - 9 3 1 3 1 1 2420-2.99 33 33 6 1 35 3 36 3 1 1 5 0 2371.30-..3 .19 6 1,' 32 47 .2 32 .2 26 23 26 IS I2 3 393.00.39 1 1 - - - is 37 33 10 16 a 336

-. 06.399 2 10 1 1 - 'S- 1

3.30-3.99 :6 13 6 9 19 8 17, 3 33 1 31.30- 2 .9 38 14 22 9 32 35 37 3 13 33 3 66' A' 3291.0 -1.99 - 23 2 36 9 16 5 33 37 14 43 5 .0 2730.30-3.99 - - 1 5 - - - - 1 14

3.30-3:99 2 - 3 - 0 3 - - I - 3 - 2Go3-3.99 Z6 30 20 37 36 29 2 9 3 is 4 9 3 3341.30-399 30 28 31 26 1 4 13 2 1 29 2 23 1! 39 335J.3 3.99 a 6 1 7 - 3 2 6 3 16 33 30 394

5.0,9 17 1 - - -- 33.30-:.99 23 1s 5 - 6 39 L9 - - 2 - 09

Z.0-.9 29 2 J7 30 3 9 3 q - 2 3 2 360i.00-31. 99 3 9 30 3 13 9 1 35 35 1 35 30 283.00-3.99 - - - - - - - 19 35 5 33 39 327

609334STRATIVE

1.30-3.99 33 361 9 76 33 - - 6 -31 273.0.0-11.99 3 6 79 10 - - 2a 60 60 0 0 3 6 396

..30..o.99 LO 2 3 i9 - --3.00-3.99 A3 -s z - 3- - 1

229

.99'

0...0-..99 6 - - 3 - - - - - - - - 6

.00- 1.99 15 .9 51 - 15 12 11 - 5 - 36

0. 00-0.9 - 1 57 3 56 56 60 4 57 30

30-.9 6 9 5 - 2 5 3 31 - -5.0-2.99 19 2.1 26 5 23 19 14 13 7 17 2 6 6 187

L.00-1.99 32 24 1 .4 32 31 3! 29 12 36 41 4, 51 .2350.00-0.9 5 - 5 .1 6 3 4 0 3 1, 1

1 3 - - -

1.00-3.99 1 - - - 12 1 - - - 27 - - - 0Z.30-2.99 ,-2 15 , 4 22 - 29 . - 021.30-L.99 5 56 27 2 3 38 1 - - 42 - 220

0.00-0.99 - 1 33 - - 59 60 60 18 60 291

AAr Trsffc Control Obe or T-chutctan

*-. 30-. 99 - - - - I I I - - A 93

3.0399 1, 1 4 1 1 5 - 3 21 - 82.20-2.99 11 12 5 7 2 2 23 26 25 17 5 15 17 1 1671.00-1.99 48 -1 51 52 12 14 16 20 31 43 25 12 31 3960.00-0.99 - 7 2 1 - I - LO 2 6 5 1 26 Z8

airm AdsIlJttjSt.ou

4.00-.,99 - - 23.00- 3.99 3 2 - - 4 -1- - - 4 - - - 132.00-2.99 9 12 3 - 26 22 L5 - 36 - I - 1231.00-1.99 42 46 56 49 30 38 45 1 -0 5 42 3 a70.00-41.99 1 - 1 ,1 - 9 62 - 55 12 57 263

S . lv Sertc*s

1.30-3.99 4 - - 6 6 8 4 - - 362.00-2.99 6 3 6 40 49 41 - - 3s - 43 313

1.00-1.99 s 13 02 .. 9 3 11 35 4 3 z1 38 17 53 353

0.00-0.99 1 5 5 25 7 2 2 78

Airo-, Air Jn..tco-

3.00-5.99 - - -

3.00-3.99 1 - - -LI 3 - 4

.3-.9 7 . - - 6 4 - - - 28 - 7 - 561.00-1.99 38 36 20 0 48 0 39 16 10 12 1 23 - 973.00-0.99 14 18 40 58 6 6 2 1 44 50 - 9 28 60 404

GKDIAL

lst.LL1.ocs a ... St ansod 1oro' icsPretst-on

3.,00-3 9 1 .. 4, - - - 1 -4.30 3.3 4 5 - - 1 1 - - 4 - - 24

11.001.79 41 .1 4 2 30 294 31 2 1 29 11 35 1 2562.00-0.99 2 a 56 59 5 1 14 58 59 10 49 24 39 '1 5

*".th.? or*C...t

5.30-5.99 2

30".. 99 - -- 23 i5 1 - 2 I 17 -

.00-3.:99 6 - - ,7 0 25 ,9 - 27 - 44

00-2.99 3 - 2 1 - - 9 - - 1431.300- 1.99 1 9 12 11 10 - - 5 6 . ~ 4

0030s. 99 - - i 9 48 - - - 55 so - 13 - .9 269

,1.00.3:99 - 3 1 2 - - 9. - 3

2. 0011.;9 10 3 1 i4 I 5 - 3 - 7Do0- .99 55 36 21 1 7 5d 9 - 1 35 2 29

0.00-0.99 L0 15 8 46 7 6 53 52 6 54 20 56 b33

230

rT

70-Dl 4D C? RC2 36 &A F75 AD ASl 2? 3A IX SLot

Envnto, ?.n~~e t -1 -Itrt i...tl, n &O I, ns.

10-21.99 1 5 - 3 10 I4I - 13 1 - S4i 2 I.39 19 16 -2 11 so 25 50 6 1 1 1 - 1 .82

1.10-.9 - - - - -

_.302.9 24 32 19 61.2 13 21 4 23 2 0 - 210J.3-1.99 1? :b 35 38 23 16 53 1 27 0 '0 . 36

.3 0.91 5 16 - -Z- 2 7 3 2 6 4 26 17 127

Deta 4 Preventiv. Dettry 7.chntcun

3.30-5.19 3- 'J. 9 13

3.-0-3.59 :3 i6 3 14 1 1 - - - 6

2. -. 9 10! )a 15 1 P3 26 2 5 31 1 1710

1.0-.9 2 9 2 57 9 26 22 6 9 30 30 16 29303.5 -- - . - 5 54 20 25 9 .3 23

Fir. Poeto

'.10-3.)9 i 7 - - - - .2.3-2. 99 25 02 22 19 32 03 32 5 5 6 11 2 5

.3 594 3 1; 29 1 1 3 22 32 5 54 .9 3 S 667.0-.9 - - I - - - - . - - - - 2 5

G;raphic.

5.30-5.99 7 1

.. 0-. 99 1 5 - I1 .30-3.99 15 0 - - 3 1 1 - - - - - - 36

2.029 7 25 2 - 4 t9 16 - - 13 6 - - 1321 0-.9 1 24 27 23 22 36 57 -331 23 12 17 258

10-0.99 - - -1 3 i - 60 60 16 31 46 43 365

Atoll Photoaranhic

5.0-5 1 1 - - - - - - - - - - 6s

30-3 9 9 9 1 6 - - - - 1I .2.0-2.9 23 26 21 - 31 12 25 8

1.30-i.99 :2 2 32 .7 11 25 2 13 9 45 .0 57 3 386

1.3.99 - . 1 1 1 1 I - Z 14

3.30-1.99 -9 33 .5 53 16 17 -2 24 27 "6 57 14 . 4820-1.99 - - - 1 - - 1 31 ' 4! .5 15 174

2.30- 2. 99 34 34 13 - 3 2 * 3 3 16 17 9Z.019 4 IS 23 26 50 30 3 1 &4 2 34 45 3 '2o 7

3 .00-11.99 - 2 16 34 3 1 ' 24 32 19 12 14 2 187

Aircraft Eldnrcal ASt-s Onl l~

I3-.9 A - - - - - 333.30-51.99 1 0 25 2 - 55 35 15 - - - 1I 142.30-2.9 23 23 16 3 23 13 L7 1;3 11 12 25 2151.Do-1.99 2 20 2 0 - 3 2 53 57 59 .o . 1; 365

3.JO-. 1 - - - - -

0.00-3.)9 26 35 27 5 23 21 2-6 2 2 '

231

-em

T

5.20-5.99 -- - - - - - -

2.20-2 .99 29 29 2 1 0 3 1 90.0-1.99 23 25 35 -3 35 3 5 2 1 35 5 3 2 3.00-0.99 3- 3- -s 35 55 3 22 --

20-.9.O9 - 29 1 -so~55 2 --

1.00-1.99 359 4 5 9 35 5 d 9 273. :099 .1 2 - 2 - 5 ~ s 2 5 4

.3 1 2 - 2 32 1 19 - j - - - .

1.00-3.99 7 5 - - 2 3 5 - - . - - - 22.0399 21 7 - .3 34 3 - -- - 5 380.0-3'99 32 29 32 53IS31O1 6 . 3 ~ - 3q009 -6 -8 .

27 4 7 '37

.3 -. 99 1 - - - - - - - - -.0399 7 29 2 1 3 - - - 3 -

1.00-1.)9 21 25 2 34 25 25 2 27 25 54 3 "' 4 9G. 00-). 99 - - a .1-, 6 32 2 2 9 2 3

232

APPENDLX F

Lettei Intioducing Survey

2:33

"'-1 .. . - 4

. .... -' .. ...~~~~~~~~~ :-- .. .-3 2' ,. . . .. ... . . . .

ar

DEPARTMENT OF THE AIR FORCEAIR FORCE HUMAN RESOURCES LABORATORY (AFSC)

WROO(S AIN FORCE BASE, TEXAS 78235

AM oil AFHRL/OR

suEjr Administrative Support for Job Requirements Study

TO: Occupational Survey Control Officer

I. As a Headquarters USAF-approved project, this Laboratory conaactsresearch on establishing job requtrements in the enlisted. specfalties.. An.important part of this effort Is the development of Job survey Instru ,entsfor use in determining minimum entry requirements into various AFSs. Inconnection with this research effort, the firm of Applied PsychologicalServices Inc., working under AF Contract F33615-78-C-0032, is required togather job requirements data at a representative sample of Air Porceinstallations,

2. We request permission for two (2) contractor personnel to vf sit yourCBPO for purposes of administering a job requirements survey to selectedsamples of enlisted personnel. The contractor personnel will bring allnecessary survey materials for group administration and will conduct thesessions under your supervision. Since the number of respondents is largeand varies by installation, we would also request administrative supportin bringing groups of personnel within your jurisdiction to a centraltesting location on a schedule that is compatible with your office. Rostersw.ill be provided or, if preferred, we will supply individual address labelsfor use in corttacting respondents through base distribution.

3, It is anticipated that most of the survey administration activity canbe conducted within a week's time. (See survey schedule at Atch 1).Further instructions and administrative material for the project will beforwarded within the next two weeks. We appreciate your assistance inthis effort.

1 AtchSurvey Schedule

Introductory letter forwarded to all Air IBas;'; included in the sample.

235

I]

A!.A093 981 APPL.IED PSYCHOLOGICAL. SERVICES IkC W*181 PA FGilPERC(PTUAL/P:ICAONMOTOAagjREINheils DASICSTOP -II :ANS* C : IE TCIA.rEC a80 a I SE L. P r.J FDRAk. M AEL r3N 1 FJI .78-.1

UNCLASSIFIED AFHRL.TQ 40.2& Oft

UPPLEMENTARY

INFORMATION

DEPARTMENT OF THE AIR FORCEAIR FORCE HUMAN RESOURCES LABORATORY (AFSC)

BROOKS AIR FORCE BASE. TEXAS 78235

RtFPLT TO

ATT OF: TSR 16 JAN 1981

SVWrCTr Removal of Export Control Statement

70. Defense Technical Information CenterAttn: DTIC/DDA (Mrs Crumbacker)Cameron StationAlexandria VA 22314

1. Please remove the Export Control Statement which erroneously appears onthe Notice Page of the reports listed . . This statement is

( intended for application to Statement B reports only.

2. Please direct any auestions to AFHRL/TSR, AUTOVON 240-3877.

FOR THE COMMANDER

WENDELL L. ANDERSON, Lt Col, USAF I AtchChief, Technical Services Division List of Reports

Cy to: AFHRL/TSE

- - . . . . . . . . . " -. . .. . . . . - : . . . . . . . .. . . - -- ., . - .- o