AD AQI+ 5 981 AIR FORCE INST OF TECH WRIGHT—PATTERSON AFU OHIO SCH——ETC FIG 5/9

PRODUCTIVITY AND JOB SATISFACTION IN RESEARCH AND DEVELOPMENT: —— ETC (

U)
SEP 71 I. .i CORBIN
UNCLASSIFIED AF ITIGSM/SM/115 2 NL

H _ _ _ _ _ _ __ _ _ _ _
‘
 —

10 28

~~ II ~
I
Hll
~35

1•1 ~
L
~
______

‘ 4 l.25
~~~~~~~~~~~~~~~
~ flffl

NATIONAL BUREAU OF STANDARDS

MICROCOPY RESOLUTIOH TEST CHART

L . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
 f r .

D DC
rN~~3 i9?;~fl
25t U1TE [J
~j~~ D~
UNITED STATES AIR FORCE
AIR UNIVERSITY

-‘ AIR
~~~ FORCE I NSTITUTE OF TECHN O LOGY
Wr ig ht-Patt .rson Air Fo rcs Ias. ,Ohio

~~
• DISTRIBUTION STATEMENT
Approved fox public release;
Distxibutior ~ Unlimited

~~~~~~~
-
~~~
 AFIT /G MS/SM/77S-2

PRODLTCTIVITY AND JOB SATISFACTION IN

RESEARCH ANT) DEVELOPME NT: ASSOCIATE D

INDIVIEUAL AND SUPERVISOR Y VARIABLES

THE SIS
AFIT/GSM/SM/77S-2 Larry J. Corbin
Capt USAF

Approved for public relea se; distributi on unlimi ted

p.

_ _ _ _ _ _ _ _ _ -~~-~~~~- -,
 -

Y~~( AFIT /GSM/SM/77S-2 /

_ _

Y I~~~~~ AND1OB~~ATLSFACTION N SE Z~~~~~~ ND

~~~~ ~~~~ ODU
[ DEVELOPMENT: ASSOCIATED INDI VIDUAL AND~~LJPE R WSOR Y VARIABLES~ j

y~’ THESIS
~~~~ (
I
_ _ _

Presented to the Faculty of the School of Engineering

of the Air Force Insti tu te of Technology

Air Uni versity

in Partial Fulfillment of the

Require ments for the Degree of

Master of Science

~~

C ED

J
IT
,.
OIST
~:~~ ~ AV LA ~~L Y
~~~~
A ,L . d
~~~ ~~~~~~~~~

~~~~~~~~~~~. c orbin 0 0 C
Gradua te Systems Mana gem~ L NOV 3 1377
L.
~~~~~~~~p—-~~~~~~~
~4JL ~~J~
Approved for public release; distributi on unlimi ted
 ACKNOW LE DGEMENTS

Historically, there seems to be a certain amount of mental pai n ,

frustra tion , and uncertainty involved in every thesis effort. I wish to

express my grati tude to the following indivi duals whose efforts during

this research lessened the impact of these fa ctors:

Captain Michael J. Stahl. As my thesis advi sor , his encourage-

ment , direction , and interest throughout this research effort proved

invaluable.

Mr. Max t~~vis, AFFDL. As my labora tory point of contact,

his assistance in obtaining da ta was most helpfu L

My pa rents, Arthu r and Be tty. Their moral support throughout

‘ this effo rt was extremely beneficial to me.

Finally, my wife , Sue. Her assistance and encouragement

during the “low points ” of this resea rch effort are extremely meaningfu l

to me and leave me indebted to her.

II
 TABLE OF CONTENTS
Page

ACKNOWLE DGEMENTS ii

~~ ‘I’ABL.ES vi

VIII

I~’rI’ B~OlJ(Jc ’rIcD N I

1.1 Overview 1
1.2 Statement of Problem Z
1.3 !...iniitations 3
1.4 Definiti ons 3
1. 5 Hypotheses 4
1.6 Objectives 5

2. LITERATURE REVIE W 6

2.1 Introduction 6
2. 2 Pr oductivity 0
2.3 Job Satisfaction 7
2.4 Leader Behavior 9
2.5 Internal Versus External Control of Reinforcement.12
2.6 Leader Reward Behavior 13
2.7 Theory Rela tionships 14
3. ~vIE1:’HODOL4Cx ~ Y 20
3.1 Introduction .20
3.2 Survey Technique 20
3.3 Sam ple .22
3.4 [)lstributlon .23
3.5 QuestIonnaire Structure 24
—

3.6 Measurement Scales 25

3.6.1 ~kitput/Productlvlty 25
3.6.2 Rotter’s I-E Scale 26
Hoppock’s Job Satisfaction Measure
3. 6. 3 27
3.6.4 Supervisory Behavior Description
( .iestlori na.lre 27
~
3.6.5 Leader Reward Behavior Instrument 28

I
l

L .L .
.. ~~~~
- - -~~~~ - :
,
.
~

TABLE OF CONTENTS
(Continued)

Page

3.7 Variable Abbreviations 28

3.8 Analytic Methods 30

3. 8. 1 Frequencies and Statistics 31

3. 8. 2 Fac tor Analysis 31
3.8. 3 Correlation Analysis 33
3. 8.4 Regression Analysis 33
3. 8. 5 Tests of Means 34
3. 8. 6 Discrimi nant Analysi s 34

4. RESULTS 36

4. 1 Overview of Analysis 36
4.2 Introduction 37

4.2.1 Deleted Var iables 37

4.2.2 Statistical Characteristics of the H
Respondents 38

4. 3 Dimensiona lity of the Measurement Scales 41

4.3.1 Rotter’s t-E Scale 41

4. 3. 2 Leader Reward Behavior In strument 43

4.4 Analysi s of Productivity Variables 45

4. 4. 1 Producti vlty Statistlcs 45

4. 4. 2 Correlations Among Output Variables 47
4. 4.3 Correlations Among Productivity and
Demographics 49
4. 4.4 CorrelatI ons Among Productivity and
Predictor Variable s 50
4. 4. 5 Tests of Means Between Produc t[ve /
Nonproductive Categories 52

4. 4. 5. 1 Tests of ....IS Ba sed on Demo-

graphic Variables 53
4.4. 5. 2 Tests of Means Based on Pre-
dictor Variable s 55

4. 4. 6 Dlscri min ant Analysi s of Productivi ty

~~~~~~~ les 60-

iv
- - - -
 - - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

TABLE OF CONTENT S
(Cont
inued)

Page
4.5 Analysis of job Satisfaction 66

4.5.1 Job SatisfactionStatistics 66

4. 5. 2 Correlations Between Job Satisfaction
and Demographics 67
4.5.3 Correlations Between Job Satisfaction
and Predictor Variables 68
4.5.4 Regression Analysis 69

4.6 Analysis of Team Development Questions 70

4.7 Summary of Results : Hypotheses Tested 74

4. 7. 1 Compa rison for Nonsupervisory

Scientists/Engi neers 74
4.7.2 Comparison for Group Leaders 79

5. DISCUSSION AND CONCLUSIONS 83

5.1 Introduction 83
5.2 Validity of Measurement Scales 84
5.3 Implications for Management of the Tested
Hypotheses 85
5.4 Implications of Team Development 88
5.5 Shortcomings of Current Research 89
5.6 Recommendat ions for Future Research 89

BIBI...IC)C~RAPHY 91
APPENDIX A: Questionnaire 95

APPENDIX B: Variable Coding Scheme 109

APPENDIX C: Correlation Matrix for Rotter Questions l

il

APPENDIX D: Factor Analysis of Rotter Questions 114

APPENDIX E: Correlation Matrix of Leader Reward

Behavlor Questions 116

APPENDIX F: Correlation Matri x for Regression Analysis

\Tariables 119
‘VI’I’A . . . 122

V
 t LIST OF TABLES

TABLE PAG E
I Output/Produc tivi ty Categories 26
II Variable Abbrev
iations 29

III Demographic Characteristics of the Respondents 39

IV Group Characteristics 40

V Rotter I-E Scale Factor Analysis (N 273) 42

Vi Eigenvalues of LRBI Factor Analysis 44

VII Leader Reward Behavior Factor Analysis 44

VIII Productivity Statistics 46

LX Correlations Among CXitputs 47

X Productivity Correlations With Demographi c 49

Variables

XI Producti vi ty Correlations With Predictor

Var
iables 51

XII Product
ive/Nonproductive 54

XIII Productive/Nonproductive Group Leaders 56

XIV Product
ive/Nonproductive 57

XV Product
ive/Nonproductive Group Leaders 59

XVI Discri minant Analysis of PVAR 1 62

X VII Dlscriminant Analysls of PVAR 2 62

XVIII Discriminant Analysis of PVAR 3 63

XIX D( scrl mlnant Analysis of PVAR 4 64

H’ vi

L
 -. -- —

LIST OF TABLES
(Continued) -;

TABLE PAGE

XX Discriminant Ana lysis of PVAR 5 64

XXI Di scri minant Analysis of PVAR 6 65

XXII Jab Satisfaction Statistics 66

XXIII Job Satisfaction Correlations With Demographic

Variables 67

XXIV Job Satisfaction Correlations With Predictor

Variable s 68

Xxv Regression Analysi s For Job Satisfaction 70

XXVI Analysis of Awareness Question 71

XX VII Analysis of Participation Question 72

vii

~~~~~~~~~~~~~~~~
.-.
~~~~~~~~~~~~~~~~~~~~~
.
~~~~~~~~~ -
.
~~~~~~~ -~~~~~~~~ --— . - - L ~~~~~~~~~ ~~.
 ~ ~~~ ~~~ ~~~~~~~~~

AFIT /GSM/SM/77S-2

ABSTR ACT

Thi s research explored the relationships of supervisory and

individual variables to the productivityand job satisfaction of scientists

and engineers in a selected Air Force research and development labora-

tory.

Productivity was defined as quantity of output and was measured

across six different types of output. Job satisfaction was measured by

the Hoppock mea sure.

Supervisory and individual variables were measured by ni ne
demographic variables and the following measurement sca les: 1)the

Supervisory Beha vior Des~~iption Questionnaire (SBDOJ, 2) the Leader

Reward Behavior Instrument (LR BI), and ‘3) The Rotte r Internal-External

(I-E) Scale of Control of Rei nforce ment. These variables were then

tested for their relationships with the criteria. -

The method was a questi onnaire administered to the indi vidual

scientist/engineer. Usable informati on was obtained on 326 civilian and

military scientists/engineers resulting in a sa mple of 272 nonsupervi sory

scientists/engineers and 54 group leaders. -.

Factor analysis indicated, with the excepti on of the Rotter I-E

scale , tha t the measurement scales were valid. The analysis of the Rotter

I-E scale indicated that the scale has questionable validity, is multi-

di mensiona l , and may be sensitive to population differences.

viii

—~ _ _ _ _ ____ ——— =
_ _ _ _ -- .--- ,— - —
- ~~~~~~~~ - - — - -~~~ - ~~~~~~~~~~~~~~~~
 — - ~~~~-.- -- .-
~~~~~~~~~~~~~~~~~~~~~ -~~~~~~ -

AFIT /GSM/SM/77S-2

For both the nonsupe rvi sory scientists/engineers and the grou p

leaders, the following relationships were found. No rela ti onship was

found between productivity and job satisfaction. Although hi gher educa-

tion , grade , and experience were associated with highe r productivity, no

single predictor variab le was shown to be significantly associated with all

six of the p roductivity variables, including leader beha vior and the Rotte r

score. However , consideration and positive leader reward behavior were

positively related to job satisfaction. Education was found to be asso-

d ated to job satisfaction : positively for the nonsupervisorv scientists /

engineers and negatively for the group leaders.

The total Rotte r score was nega ti vely associated with job sa tis-

faction for the nonsupervi sory scientists /engineers.

ix

-
_ _ _ _ _ _ _ _ _ _ _ _ _ _
. ~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~ -
 PROIXJCTWITY AND JOB SATISFACTION IN
RESEARCH AND DE VELOPMENT:
ASSOCIATED INDIVI DUAL AND SUPERVISOR Y VAR IAB LE S

CHAPTER 1. INTRODUCTION

1. 1 Overview

The questions of orga nizati ona l (situa ti onal) determinants and!

or individual determinants of ou tpu t and job satisfaction as they relate

to reward perceptions in a Researc h and Development (R&D) Organi 7a -

don have been only recently investigated. The major portion of reward-

contingency research has been directed at the impact of monetary rewa rd s

on motivation and performance (Si ms and Szilagyi , 1975). Addi tionally ,

such topics as supervisory style , morale , employee satisfaction , c orn-

munication s systems, etc. , have been studied for a number of years

although these topics have not been thoroughly rela ted to rewa rd-contingenc y

research. However, the current gove rn ment position on R&D is to get

more R&D for the dollars expended. Vincent (1972) identified thi s change

in R&D policy and recommended that “in order to get more R&D for the

dollars expended, it becomes impe ra tive that ma nagement be able to iden-

ti~.fy factors which influence productivity”.

The pu rpose of this research then is to investi ga te the rela tion-

ships among indi vidua l perceptions of rewa rd -

both as a genera li zed
expectancy and as a perception of leader reward behavior, and leader

behavior (supervisory style) as they relate to productivity and job satis-

faction.

I
 1. 2 Statement of the Problem

What is the relationship between the productivity and job satis-

faction of AF R&D scientists /engineers and their generali zed expectancy

of reward? Is the productivity and job satisfaction of scientists /engineers

related to their perception of leader rewa rd behavior? Does a relation-

ship exist between the individual’s generalized expec tancy of reward and

the individual’ s perception of leader reward behavior? What effect does

the supervisory style have on productivi ty, j ob satisfaction , and the indi-

vidual’s perception of leader reward behavior?

The ba sic prob lem can be related to the Lewinian field theory

whereby the work behavior of the indi vidual was postula ted to be a func-

tion of the ind ividua l and his environment , i. e., Behavi or = f (indi vidual

X environment) (McCa rrey and Edwa rds , 1973). Thus, productivityand

job satisfaction can be thought of as being determined b y both the indi vi-

dual and the envi ronment. The supervisor is the individual’ s main

contact with the envi ronment : the supervisor controls the positive or

negati ve rewards which are given to the individual. Similarly, the

individual’s expectancyof reward, both as a generali7ed and leader-

oriented concept i mpacts upon productivity and job satisfaction. The

goa l then is to de termine how certain management practices affect the

individual’ s perceptions and ultima tely productivityand job satisfaction.

L ~~~ .- -
~~~~~ ~~~~~~~~~~~~~~~~~~~~~ .~~~~~~~ - ~
-.- .
 1.3 Limitations

This study investigated AF R&D scientists /engineers only. It

was further limited to one AF laboratory and i nvolved working level

scientists/engineers and group leaders. This policy included both

military and civilian personnel , but exclude d all levels of management,

staff and administrative personnel, technicians, and trainees except for

the group leaders.

1.4 Definitions

. The following definitions serve to clarify the terms which will

be used throughout this report:

1. Internal: an individual who perceive s positive and , ’or ne ga-

tive events as bc ing a consequence of his own actions and thereby u nde r

persona l control.

2. External: an individual who percei ves positive and /or nega-

ti ve events as being unrelated to his own behavior in cert ain si tuations

and therefo re beyond persona l control.

3. Initiating Struc ture (S): Reflects the extent to which an

individual is likely to defi ne and structure his role and those of his

subordi nates toward goa l attainment. A high score on this dimension

chara cterizes individuals who play a more active role in directing

group activities through planning communicating informa tion, trying

out new ideas , etc.

- “ ~~~~~~
.
~~~~~~~~~~~~~~~~~~~~~~
. - ~~~~ -~~ -- - .~~-— ~~~~~~~ m ~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~
 4. Consideration (C): Reflects the extent to which an individual

i s likely to have job relationships cha rac teri zed by mutual trust, respect

for subordinates’ ideas and consideration of their feelings. A high score

is indicative of a climate of good rapport and two-way communication.

A low score indicates the supervisor is likely to be more i mpersonal in

his relations with group members.

5. Productive: an indi vidual, who, based on the productivity

measures used in this research effort , produced one or more items of

output.

6. Nonproductive: an individual who, based on the product

ivity

measures used in this resea rch effort , did not produce any output in a

pa rticular category.

1.5 Hypot
heses

The following hypotheses were tested:

1. A positive relationship exists between productivityand job

sat
isfaction.

2. A positive relationship exists between producti vi ty and an

internal.

3. A positive relationship exists between job sa ti sfaction an’!

an internal.

4 -
 4. A positive relationship exists between both productivity and

job satisfactionand positive leader reward behavior. A negative rela-

tionship exists between both productivity and job satisfaction and pu ni-

tive leader rewa rd behavior.

5. A positive relation ship exists between both p roduc tivi ty and

job satisfaction and the supervisory style of “Consideration” .

6. A negative relationship exists between both productivi ty and

job satisfaction and the supervi sory style of “Initia ting Structu re” .

1.6 C*jectives

The following objectives will serve as guidance during this

researc h effort:

1. Using the de veloped survey, collec t da ta on a samp le of

scientists/engineers from an AF R&D laborato ry .

2. Test the rela tionships a mong the following factors: produc-

ti vi ty , job satisfaction, generalized expectancies for internal versu s

external control of reinforcement , leader rewa rd beha vior, super-

isory style, and certain statisticalcontrol data such as education,

v

age, exper
ience, etc.

3. By using the collected data, attempt to recommend certain

laboratory policies which would improve/maintain the curren t ma nage-

ment practices.

5
 CHAPTER 2. LITERATURE REVIEW

2.1 Introduct
ion

There are many basic problems when one tries to study an area

such as productivity. Fi rst, there are many methods which may be

used to measure productivity. These various methods have certain

advantages dependent upon the circumstances. One must decide what

method best fits the situa tion dependent upon the desired result of the

measurement and the individuals being surveyed.

The second proble m deals with both the organi zationa l and m di-

vi dua l diffe rences which i mpact upon productivity. There are a variety
of factors suc h as supe rvisory style, innova tion , morale , sa tisfaction ,

etc. , which migh t affect the productivity of individuals and ultimately.

the orga niza ti on. By necessity, these fa ctors must be limi ted to allow

for a mana geable re searc h effort. In particular, the following factors

will be investi gated in this research effort: supe rvisory style (leader

behavior), internal versus external control of reinforcement, and

leader reward beha vior. Prior to reviewing the li tera tu re rega rdi ng

these factors, the concepts of productivity and j ob satisfaction must be

described as they relate to measures of these concepts.

2. 2 Produc ti yi~y

Many forms of identifying productivity are available. McCa rrey

and Edwards (1973) list nine performance measures which are

6
 dimensions of scientist performance: 1) productivity rank -
order
scores (peers), 2) creativity rank -
order scores (
peers), 3) depart-
mental percentilescore, 4)professional recognition score, 5) comm-

unica tions rank -

order scores (peers), 6) quality of published work

(peer ratings), 7) origi na li ty of published work (peer ratings), 8) cita-

tion rate (number per paper), and 9) publications rate (number per

year). for purpose s of this resea rch, however, only output tha t can be

quantitatively identified will be used to represent a worker ’s productivity.

Although a qualitative assessment of R&D productivity is ofte n more prac-

tical than a quantitative measure , quanti ty is easier to determine and

measure, and much less subjec ti ve (Hughes Study , 1974).

Due to the practicality of a quantitative measure and the ease

with data may be ga thered , a quantitative producti vi ty measure was used

in this study. The specific details of the productivity measure are des-

cribed in the following chapter.

2.3 J~b Satisfaction

The basic factors of job satisfaction may be broken into two types:

those characteristics of the individual and those characteristicsof the

job. As factors associated with job satisfaction, these factors incor-

porate the personal characteristics of each indivi dua l and the organiza-

ti ona l characteristics of each job.

.
~

-- ——
-
—
..—
- — . -~
 In particular, the individua l characteristics of age , education,

intelligence, sex, occupa tional level and individual difference s have

been investigated as they relate to job satisfaction (Fournet , 1966).

Various research efforts have concluded that each of these character-

istics can have considerable i mpac t on an indllvidual~~ job satisfaction.

For exa mple, some research indicates tha t satisfaction or morale

increases as the job level progresses (Fournet, 1966). This research

F
indicates tha t a Ma slow -type satisfaction of different needs may be pre-

sent which progression from lower job levels to higher job levels

sa tisfies.

In contrast to individual chara cteristics, many studies have

indicated that the work situation has tremendous i mpact on job satis-

faction. Some factors which ha ve been related to job satisfaction include

organization and management , immediate supe rvi sion , social environ-

ment , communication , securi ty , monotony, and pay. Job sa tisfac tion

could de pend , in part, upon the effect of these fac tors by the extent to

which they lower or increase the worker ’s self-evaluation (Vroom. 1962).

- In an atte mpt to capture the individual ’s job satisfaction , Hoppock

in 1935 formulated a four question measure for job satisfaction. The

intent of the four questions was to capture variou s aspects of an indi-

vidual’s satisfaction with his job in a globa l rather than specific measure

of job satisfaction.

_ _ _ _
- --
~ ~~~~~~~~~
_ _ _ _ _
 The validity and reliability of this measure has been fairly well

established. Using a data base of over 28, 000 responses. McNichols,

Stahl , and Manley (1976) found that Hoppock’s measure appears to be

a meaningful global measure of an employee ’s perception of his job

satisfaction over a wide range of orga ni 7ations. In particular, it was

found that ”. ., the measure performs consistently when applied to a

variety of sample populations including many different job categories,

organiza tiona l levels and demographic groupings”.

Thus, the Hoppock measure of job satisfaction was used in thi s

study. The specific details of the measure are described in Chapter 3.

2. 4 Leader Beha vior

Among some of the large scale psychological resea rch prog rams

on leade r behavior, one which had considerable i mpact was conduc ted

at Ohio State University during the period 1945-1956. This research

revea led two basic dimensions of leader behavior in the forma l organi-

za tion, “Consideration” and “Initiating Structu re” . These basic

dimensions were derived from the initial investigati on of eleven di m-

ensions of leader behavior descriptions (Stogdill and Coons, 1957).

As a result of the research effort , it was found that leader behavior

could be described in the di mensions of “Consideration” and “Initiating

Structure”.

_ _ _ _ _ _ _ _ _ _
 The leader behav
ior research resulted in three separate devices
to measure the behav
ior: the Leader Beha vior Description Question-

na ire , the Superv

isory Beha vi or Description Questionnaire, and the

Leadership Opinion Questionnai re. All questionnaires re ly upon the

“Considera tion ” and “Initiating Structure” concepts, but approach the

measurement of leader behavior from di fferent avenues.

The Leader Behavior Description Questionna i re (LBDQ) is a

measure ment of the subordinate ’s perception of leader behavior on-

ginally containing 150 questions. The Supervisory Behavior Description

Questionnaire (SBDQ) was developed by Fleishma n con taining 48 ques-

tions for use in i ndustry .

A recent study by Szi lagyi and Kelle r (1976) has shown the

SBDQ to be a “bette r measure ” than the LBDQ by using concurrent vali-

di ty analysis. Although no significant differences were found in corre-

lat
ion with the leader consideration dimension , there were signi ficantly

di fferent results reported in the leader i n i t i a t i n g structure dimension

ith the SBDQ being more valid.

w

In contrast to the LBDQ and SBDQ , the Leadership Opinion Ques-

tionnaire (LOQ) approaches the measurement from a diffe rent perspec-

tive. The LOQ is a Ukert-type attitude scale which assesses leadership

attitudes. Each supervisor is asked not to indicate how he performs

in the work situa tion but how he believe s he should perform (Stogdill

and Coons, 1957).

tO

~~~~~~ — ~- -.a~---— — .S _ . -
_- —
_ _
---- — -~~ _ —_-
_ _- .— —
—. _ _ — .
~~~~
 __________________

The design of all questionnaires is to provide independent rnea s-

ures of “Consideration” and “initiating Structure”. The scales for

these two dimensions have been shown to be reliable and independent in

a wide variety of si tua tions (Stogdill and Coons, 1957). Thus, due to

the rating of the SBDQ as a “better measure” , it will be used in this

study. Additionally, however , there is yet little evidence of the predic-

tive validity of these two di mensions nor the kinds of situa tional variables

which might effect such validi ty (Korma n , 1966).

In an atte mpt to capture the si tuational variables which affect the

predictive validity of these di mensions, Kerr found tha t the variables

which moderate the validity of the lea der behavior prediction are the fol-

lowing: subordinate need for i nformation, job level , subordi na te expec-

ta tions of leader beha vior, perceived organizational independence, leader’s

similarity of atti tudes and behavior to managerialstyle of higher manage-

ment, leader upward i nfluence, and the characteristics of the task (Kerr ,

1974). Thus, the relationship between leader behavi or , employee atti-

tudes , and jo b performance is substantially affected by the organi 7ation

in which the leadership takes place.

Since the supervisory style has considerable i mpac t on produc-

rivity and job satisfac tion, this concept was investigated using the SBDQ.

The specific details of this measure are discussed in the followi ng

chapter.

11

—— ~~~~-~~~_~- .- - -.- _ _ -
~
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

2.5 Internal Versus External Control of Reinforcement

Various researc h efforts ha ve been conduc ted which have con-

cerned themselves with man ’s ability to control his personal environment.

In particular, the role of reinforcement, rewa rd , or gratification is

recognized as a crucial one in the acquisition and performance of skills

and knowledge (Rotter , 1966). One of the resea rch efforts in this

regard is the Inte rnal-Externa l Control of Rei nforcement described by

Julian B. Rotter.

Rotter ’s theory is based on a genera l principle that: “Internal

control refers to the perception of positive and /or nega tive events as

being a consequence of one’s own actions and thereby unde r personal

control; external control refers to the perception of positive and/o r

negative events as being unrelated to one ’ s own behavior in certain situ-

ations and therefore beyond personal con trol” (Lefc ourt, 1966). Thus,

the basic theoretical background for Rotte r ’s Inte rnal-External concept

is that a reinforcement acts to strengthe n an expectancy that a parti-

cular beha vior or event will be followed by that reinforcement in the

future. If the reinforcement is perceived to be contingent upon the

individual’s own behavior , reward or punishment in the fu tu re will be

seen as causally related to past behavior. Ukewise, if the individua l

senses no relationship between behavior and rewa rd or punishment. the

rewa rd or punishment will be perceived to be beyond the individual’s

control. Thus, Rotte r (1966) hypothesIzed that “depending upon the

12

L _ _ _ _ L
~~~
2”
~~~~~~.
 fl
~
individual’s history of reinforcement, individuals would di ffe r in the

degree to which they attributed reinforcements to be their own actions ”.

The items in the Internal-External (I-E) scale deal exclusivel y

with the individual’s belief about the nature of the world, that is, the

items are concerned with the subject ’ s expectations about how rein-

forcement is controlled.

In general, most studies have reported an unusually consistent

set of fi ndings. The validity and reliability of the I-E scale has been

extensively tested and on the whole , substantiated the concept ’s useful-

mess in severa l areas. Data tends to support Rotte r ’s contention that

the internal-externa l control theory is a genera li 7ed expec tancy opera-

ting across many si tuations (Joe , 1971) .

Due to the possible impact of the feelings of power/powerlessness

which are entailed in the I-E concept, this concept was investigated in

the study. The specific details of the I-E scale are described in

Chapter 3.

2. 6 Leader Rewa rd Behavior

In 1971, W. E. Scott , A. J. Reitz, and R. D. Johnson developed an

instrument which is designed to measure the degree to which the sub-

ordinate perceives that the rewards, or outcomes (positive or nega tive),

he receives through his supervi sor reflect his performa nce, or accom-

plishments , on the job . The instrument is based on two di mensions:

13
~
—. . — - . - ---
~~~~ — ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

“positive reward behavior ”, representing the relationship between good

performance and leader-administered rewards such as recognition and

pay i ncreases, and “pun

itive reward behavior”, representing the rela-
tionship between low job performance and leader corrective behavior

(Johnson , 1973).

In a study conducted in a major midwestern universi ty ’ s medical

center using the instru ment, it was found that across all occupational

skills, there was a relationship between job satisfaction and positive

leader rewa rd behavior. However, in relation to j ob satisfaction and

punitive reward behavior , varying relationships existed ac ross occupa-

tional skills (Sims and Szila gyi , 1975).

The effects of leader reward behavior were found by Stevens (1976)

to have an associati on with producti vi ty in an AF research and develop-

ment laboratory. For this rea son , the Leader Reward Beha vior instru-
ment was used in this study and its specific details are described in

Chapter 3.

2. 7 Theory Relationships

- One commonly held view of supervision for R&D engineers is tha t

the supervisor should do little except keep out of the way of his subord i-

na tes. This viewpoint indica tes that the soundest way to encourage hi gh

achievement is to secu re good people, give them good equipment, and

then leave them alone. However, current results show that this view-

point may not be entirely accura te.

14

-. - -— - -— - - -- -~~~~~~~~~~ “ - i~~~~ — ~~ -~~~~ _ -

~~~~
 rn a research effort conduc ted by Pelz, it was found that producti-
v
ity was highest when independence frOm the supervisor is combined with

ith him , that is , when the individual has frequent i nter-

frequent contact w

action with the supe rvisor , but also has considerable voice in the fi na l

decision s (Pelz, 1956). Additiona l support for such a theory is fro m

Lawler and Hall who conclude that the more the engineer feels he has

control over what goe s on, the more he feels his job allows him to be

creative and satisfying (Lawler and Hall , 1967).

In light of these conclusions, wha t would one expec t from a survey

of R&D engineers using the SBDQ? In general (placing aside individual

di ffe rences between engineers), one would expect that R&D engineers in

a labora tory would percei ve their supervisor as being ra ther low or mode-

ra te in “Initi a ting Structu re” and high in “Considera tion”. Thus, the
supervisor would provide an environment cha racteri zed by mutual trust,

respect for his subordina tes’ ideas, and consideration of thei r feelings,

while raking a less active role in structuring the roles of his subordi na te s

toward goa l attainment. One would expect an unsuccessful supe rvi sor to

exhibit or be percei ved by his subordinates as exhibiting behavior contrary

to that described above , This point is consistent with Stogdill and Coons

(1957), wi th Hill and Hunt (1973), and wi th Bradwhat (1970), whose data

showed respondent preference for structu re to di minish at highe r orga ni-

za tiona l job levels. Addi tionally, Bradsha w (1970) found preference for

consideration to be relatively con stant across different organiza tiona l

levels.

15

,
. - -.
_ _ _ _
~ - . -—
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
. - . .

However, could the individual’s perception of his supervisor be

based on some other fa ctor suc h as the individual’s expectanc y of inte r-

nal versus exte rna l control of reinforcement or the leader ’ s percei ved

rewa rd behavior? This author believes tha t the individual’ s self-

conception would i mpact greatly on his perception of his supervi sor.

Additional studies tend to support this expec tation . The individual who

perceives tha t he has control over what happens to him may conform or

may go along with suggestions when he chooses to and when he is given

a Con sciou s alterna tive. However, as indicated in one study, if such

suggestions or attempts at manipulation are not to his benefit or if he

perceives the m as subtle atte mpts to influence him without his awa re-

ness, he reacts resisti vely (Rotter, 1966) . Thus, if a supervi sor

were attempting to slightl y structure the work situation , the internal

may perceive the supervisor in a negative sense of “Ini tiating Structure”.

Addi tionall y, Evans (1973) found tha t peop le with a high inte rnal orien-
ta tion (on the I-E scale) have stronger consideration relationships than

do people whose orientation is exte rnal.

-
In this sa me vein , one could hypothesize tha t most engineers

should express internal tendencies. If this hypothesis is correct , the

ser
ies of studies provided by Rotter (1966) express strong support for

the hypotheses that an internal is likely to:

(a) be more alert to those aspects of the envi ronment which pro-
vl ded useful i nformat
ion for hi s fu ture beha vior:

16

.- .~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~
 (b) take steps to improve his environmental condition ;

(c) place greater value on skill or achieve ment reinforcements

and be generally more concerned with his ability , p articularly his fail-
ures; and

(d) be resistive to subtle atte mpts to influence him.

If these hypotheses are valid, the implications for successful leade r

behavior in an R&D labora tory are evident , assuming most individuals

are i nternals.

Additi onally, as suggested by the work of Sims and Szila gyi (1975),

the interna l who has the desire to con trol his future will probabl y react

fa vorably to a supervisor who he perceives to be high in positive reward

beha vior. As stated by Lawler and Porter (1967), “

... the greater the
va lue of a set of rewards and the higher the probability tha t receiving

each of these rewards depends upon effort , the greater the effort that

will be put for th in a given situation”. The subordinate’ s perception of

his chance of receiving rewards should i mpact both on job satisfaction

and productivity . Thus, the combine d effect of leader behavior, inte rnal-

external theory, and leader reward behavior for R&D engineers are fa r-

reaching.

If most engineers are internals, they would show more overt

striving for achievement tha n externals who feel tha t they have littl e

control over thei r reward s and puni shment. This hypothesis has been
substantiated and internals tend to manifest greater interest and effort

17

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 -— - -

I
---

1
-- . - .- —
P.-
~~~~~

in achi evement-related activities than do externals (Lefcourt, 1966).

In some respects , Rot ter ’s inte rna l-external concept and the leader

rewa rd behavior concept cou ld be compared to a theory advanced by

Vroom of ego involvement. In this theory , the degree of job involvement

(an appa rent correlation between involve ment and leader beha vior can be

drawn ) for a particular person was measured by his choice of “ego ”

ra ther than extrinsic factors in describing sources of satisfaction and

di ssa tisfaction on the j ob. A person was described as ego-involved in

a task or job to whatever extent his self-esteem (or internal character-

istics) is affecte d by his perceived level of performance (Vroom, 1962).

Vroo m found tha t persons high in ego-involve ment were ra ted hig her in

job performance than those with low ego-involve ment. Thus, one could

hypothesize that an internal who perceives his rewards as based on his

own act
ions would tend to be more ego-involved in a work setting.

Vroom (1962) discove red that individuals ego-involved in their job s

were more affec ted by the extent to which their jobs ga ve the m an oppor-

tunity for self-expression and tha t the grea ter the opportunity for self-

expression, the greater the job satisfaction and productivity.

Additionally, studies have shown tha t job satisfaction and produc-

tivity is congruent with the immediate supervisor’s persona l support and

interest for the worker (Fourmet , 1966). This result would tend to sup-

port the hypotheses tha t in an R&D laboratory , the successfu l su pervi sor

would be ra ted hig h on the “Consideration” dimension.

l8

-.—- .~~~
~
-- -- —-- -

Thus, it is clear from the previous studies tha t producti vity and

job sat
isfaction might be affected by such factors as supervisory beha vior,

leader reward behavior, and the individual’s perception of his control of

reinforcement. The methodswhich were used to identify the relation-

ships among these factors are described in Chapter 3.

19

- --
----
. ~~~~~~~
— - -— - . .-
~~
 — .-
- — ~~~~-- - -
~~~

CHAPTER 3. METHODOLOG Y

3. 1 Introduction

The purpose of thi s chapter is to 1) discuss the survey technique

which will be used, 2) describe the sa np le, 3) describe the distribution

method, 4) describe the questionnaire’s structure , 5) desc ribe the meas-

urement scales, 6) list the variable abbreviations, and 7) discuss the

analytic methods which will be used to analyze the data .

3. 2 Survey Technique

The use of a questi onnaire as a data collection techniaue is fairly

commonplace in the study of productivity and job satisfaction. Addition-

ally, the need for a large volume of data in a relatively short peri od of

ti me was the pri ma ry factor involved in selecting the questionnaire over

other da ta col lection techniques. There were 540 scientists /engi neers

selec ted to participate in thi s survey and they were allowed four weeks

to comp lete and retu rn the questionnaire.

Due to the nature of some questions , the question of anonymity

was satisfied by the survey technique . Persona l data on each respon-

dent was obtained by some demographic questions which did not allow

for identification of specific respondents. Additionally, the response

rate and validity of responses tend to be higher when anonymity is

insured for the respondents.

20

-I
 Some of the inherent disadvantages of the survey technique are

response ra te and the respondent’s psychological feelings about involve-

ment. Each of these disadvantages was pa rtially compensated for by

the location of the survey partic ipants. Additional action was ta ken to

further compensate for these disad vantages.

To insure a response ra te which would p ermit a meaningful

analysis of the data, the comple xity of the questionnaire was reduced as

much as possible , although thi s factor could still limit the number of

responses. Additionally, the Commander of the labora tory surveyed

was personally inte rested in this study and signed a cover letter which

w as attached to the questionnaire requesting each respondent’s assist-

ance in supporting this research effor t. Appendi x A contains these

letters.

The second disadvantage of the survey technique was partially

modified by the i nvolvement of top management in this research effort.

Besides providing a charge number for the time involved in completing

he survey, the Commander ’s letter promised a short summa ry of the

t

~urvey results to all interested participants. The questionnaire also

contained a letter from the researcher stating tha t each participant was

welcome to contact the researcher on any questions tha t migh t arise.

Addi tionally, space was provided in the questionnaire for comments

concerning the scope, intent, and length of the questionnaire.

21

• .~~
- - - .
~ - - .- - .
--- ~~~~ -
.. .. .

To insure that the questionnaire was returned in time for appro-

priate analysis, a deadline was imposed on the respondent. A da te was

identified which allowed the participants four weeks to retu rn the corn-

pleted questionnaires. It was felt that this time period was long enou gh

to include most periods of worker absenteeism, including vacations.

3.3 Sample

Support for this research was provided by the Air Force Flight

Dyna mics Laboratory (AFFDL) at Wright-Patterson Air Force Base,

Ohio. As previously indicated, laboratory mana gement actively sup-

ported this research. In addition, a sufficient number of scientists !

engineers existed within this laboratory to provide a good data base.

The A FFDL administrative offices provided the necessary per-

sonnel i nformation to identify specific individuals to receive the survey

quest
ionnaire.

Since the intent of the research was to survey primarily first-

line supervisors and nonsupervisory scientists/engineers, 540 scien-

tists/engineers were identi fied out of approximately 1000 people working

in the laboratory. These individuals were identified by using laboratory

personnel ros ters and inc luded all individuals which meet the above

survey criteria.

22

---- _ .
3.4 Distribution

On 8 April 1977 , 540 questionnaires were distributed to the fi ve

divisions within AF F DL. By 22 Apri l 1977 , the response rate was 53

percent and gradually leveling off . By the c ut-off da te , I May 1977 , 61. 1

percent of the questionnaires had been re turned.

Fou r of the returned questionnaires were deleted from the survey

ana lysis due to the respondents failure to comple te any of the measures.

These deletions resulted in 326 questionnaires (60. 4 percent of the sur-

vey samp le ) being included in the final data anal ysis.

The 326 questi ori~iaires consisted of 273 questionnaire s with no

missing da ta and 53 questionna i res wi th some missing da ta . The missing

da ta was generally restricted to one of the parts, basicall y the Rotter

I-E scale, for any one respondent and provi ded usefu l data in the ana l-

ysis of th~ -ether completed parts. The i mpact of excluding the 53

questionr.aires was considered more critical tha n the impact of the

issing data on the analysis.

m

Due to the missing values, some analys i s i s based on a sample

size of 326 while some analysis is based on a sample si7e as low as 214.

Additionally, the group leaders accounted for 54 of the returned ques-

ionnaires.
t The sample si7e which is pertinent for a particular analysis

is included in the discussion of the results.

23

a — - --- - - - - -- - -
- ~~~~~~~~~~~~~~
~ --- ---.-— - .-- -,- - - -- - — , -~~~ . - - .- — .-- -. -
~~~~~~~~~~~~~~~~

3. 5 ~~iesti onnaire Structure_

The questionnaire used in thi s study consi sted of five previously

va lidated and reliable instru ments. The first draft of the question-

naire consisted of the following sections: output/productivity and dem o-

graphic questions, Rotter ’s I-E scale , Hoppock’ s Job Sa ti sfac tion measure,

and the Supervisory Behavior Description questionnaire.

Further analysis and discussion of this draft resulted in the modi-

fication of some of the statistical and demographic questions and the

addition of two sections dealing with leader reward behavior and a man-

age ment develop ment progra m currentl y being i mple mented in the

laboratory at the request of the laboratory management.

The fina l questionnaire consisted of five main part s plu s two

cover letters (see Appendix A). Parts A, B. and C we re similar to

the draft questionnaire and consisted of the following question ca tegories:

output/productivity and de mographic informati on, Rotte r’s I-E sca le , and

Hoppock ’s Job Satisfactionmeasure.

Part D was modified to include both the Supervisory Behavior

Description questionnaire and the added Leader Reward Beha vior instru-

ment. This p art required the respondent to identify his perceptions of

his supervisor’s management behavior and his supervisor’s reward

behavior.

Part E was added at the request of laboratory management as a

means to determine the respondent’s awareness of the team development

progra m currently being conduc ted in the laboratory .

24
 Due to the proven validity and reliabili ty of the instruments used

in the questionnaire, no pretest of the questions readability and compre-

hensibility was felt to be necessary.

The derails of these measurement scales are discussed in Section

3.6.

3.6 Measurement Scales

Several measurement scales were used throughout this question-

naire , basicallyøorresponding to the different instruments which were

contained within. The purpose of this section is to describe the various

scales used in the questionnaire. The measurement scales which are des-

cribed in turn are :

1) Output/productiv ity;

2) Rotter’s I-F scale;

3) Hoppock’ s Job Satisfaction measure:

4) the Supervisory Behavior Description Questionnaire; and

5) the Leader Reward Behavior Instru men t.

3. 6. 1 Output /Producti vity

-
The ou tpiu mea surement was a quantitati ve mea sure only and

was si milar to the measure used at AFFDL by Stevens (1976). As

indicated in Table I, the productivity of each respondent is determi ned

by the identification of the quantity of each output ca tegory over the past

two years. All outp uts are t rea ted separately in any analysis.

25
 TABLE I
Output/Productivity Categories

Variable
Number Output Included

1 Published Papers in Technica l or Professional Journals

2 Unpublished Manusc ripts

3 Technical Reports

4 Technical Memorandums

5 New or i mproved processes, products, and techniques,

and patents or patent application.

6 Hardware/software specifications, test reports. test

plans, statements of work , requests for proposal

7 Oral presentations to technical or professiona l audiences

The output/productivity section of the questionnaire can be refer-

red to in Appendix A , Part A , Section 1.

3.6.2 Rotter’s I-E Scale

Rotter’s I-F scale is a forced-choice, 29 item battery consisting

of 6 fillers and 23 ques tions wh ich are summed for a total score. The

~~a1e is designed to find out the way in which certain i mportant events in
our society affect different people. As a measure of personal belief,

here are no right or wrong answers to the Rotter questions, however, a

t

low score indicates tha t the person feels he has control over his rewards

while a high score indicates that the person feels he has little control

over his rewards.

26

.- . -------~~ -~ ----~~ - — .-
- - - .-~~~~~~~~ -~~~~~
 _

The I-F scale , containe d in Part B of the questionnaire, was deter-

mined by Rotte r (1966) to be one dimensional. Additi ona l stu dies, h ow-
ever , by Cherlin and Bouque (1974 ) and Guri n (1969) ha ve indicated that

the scale is multidi mensional. The ana lysis of this question will be
di scussed in the section on analytic methods.

3. 6. 3 Hoppock’ s Job Sati sfaction Measure

Hoppock’ s job satisfaction measure is a four-question surve y

which deals with the respondent ’s perceptions of his job . Using a seven

point Li kert-type scale , the answers to each question are scored from 1

to 7. The indi vidual answers are summed to a rri ve at a coral score whose

value then ran ges between 4 and 28, with the lowe r score s indicating 1es
~
job satisfaction.

The Hoppock measure is Part C of the questionnaire. It has been

shown to be a globa l measure of job satisfaction which is re la ti vely easy

to administer and is highl y vali d and reliable (Mc Nichols, Stahl , and

Manley . 1976).

3. 6. 4 Supervisory Behavior Description Questionnaire (SBDQ

~
The SBDQ, contained in Part D of the questionnai re , was developed

by Edwi n A. Fleishma n and was based on the “Consideration” and “Init-

iating Structu re ” for ma t re sulting fro m the Ohio State studies in the

1950s. The SBDQ is a 48 ite m que stionnaire where each item is scored

from 0 to 4, dependent upon the re sponse. The “Considera ti on ” portion

(SBDQC) of the quest
ionnaire Consists of 28 questions for a range of

27

.I —-- -. -- -- -“ -------- - - --- — -- - - -—-- —

~ .
--- -
- - - -.-

possible scores of 0 to 112, while the “Initiating Structure ” portion

(SBDQI) consi sts of 20 questions for a range of possible scores of 0 to

80.

Evidence tha t each portion of the SBDQ is valid and reliable was

presented by Fleishma n (1957) where most of the items assigned to each p

porti on had high loadings with tha t dimension and insignificant loading

with the other di mension.

3. 6. 5 Leader Rewa rd Behavior Instrument

The Leader Reward Behavior Instru ment, contained in Part D of

the questionnaire . is a 22 ite m instru ment dea ling with a subordina te’s

perception of his supervisor ’s use of a reward syste m. Deve loped by

Scott , Reitz, and Johnson in 1971, the Leader Reward Behavior Instrument ,

which utilizes a seven-point Likert scale , consists of two portions: the

Leader Reward Behavior Positive (LRBP) and the Leader Reward Beha vior

Punitive (LRBN). The LRBP consists of 16 questions for a total range

of possible score s of 16 to 112, whi le the LRBN consi sts of 6 questions

with a ra nge of possible scores of 6 to 42.

The technique of factor analysis is used to test the dimen sion-

ali ty of thi s instru ment and is described in Section 3. 8.

3.7 Var
iable Abbreviations

Table II lists the different sections identified in the ques tionnai re

and thei r associa ted variables and variable abbreviations used for anal-

ysis purposes.

28 ~~~~~

. - -- --- . - .- _ _ _
 TABLE II
Variable Abbre via tions

Variab le
Category and
Abbreviation Variable Title

Demographic Control Variables

AGE Age
GRADE Civilian grade or military rank
YRRSEC Actua l years in section
YRSIS Actua l years under immediate supervisor
EDCTN Level of education
NCWR Percent of current work in researc h
NCWD Percent of current work in develop ment
CONM Percent of current work in contract monitoring
YEXP Years of scienti fic /engineering experience

Pr oductivity

PVAR 1 Publishe d papers in professiona l ‘technical jou rnals

PVAR 2 Unpublished manuscripts
PV AR 3 Technical Reports
PV AR 4 Technical Memorandums
PVAR 5 New or improved processes, products. and techniques,
and patents or pa ten t application
PVAR 6 Ha rd wa re/softwa re specifications, test reports . test
plans , statements of work , requests for proposals
PVAR 7 Oral presentations to technical or professional audiences

Group Leade r

GPL Identification of group leaders

I-F Scale

Ri to R29 Rotte r ’s 29 questions

ROT Rotter ’s total summed score

29
 _ _ — - .
_ _ _ _ _ _ _ _ _ _ _ ~~~~~~~

TABLE II
(continued)

Variable
Category and
Abbrevi a ti on Variable Title

Job Satisfaction

Hi to H4 Respondent ’ s score for each Hoppoc k question

HOP Job satisfaction summed score

Supervisory Behavior

Si to S48 Respondent ’s score for each SBDQ questi on

SBDQC Conside ration t~ota l score
SBDQI Initiating Structure tota l score

Leader Reward Behavior

Li to L22 Respondent ’s score for each lea de r rewa rd behavior

questi on
LRBP Positive re wa rd behavior total score
LRBN Punitive rewa rd behavior tota l score

Tea m Development

AWARE Awareness of tea m development progra m

PART Pa rticipation in team develop me nt progra m

MILC IV Identificati on of Milita ry Respondents

3. 8 Ana lytic Methods

The objectives which were introduced in Chapter 1 dic tate the

use of certain analytic techniques.

Many of the variable relationships require relatively simple stat-

istical techniques while others require more complex and detailed

analytical techniques. The followi ng ana lytical techniques were iden-

tified for application to the survey data. In additi on , the technique s

30
-
k. .
~~
 were available in local computer facilities through the Statistical Package

for the Social Sciences (SPSS).

Due to the fact tha t the group leaders are inc luded in survey da ta ,

analysis of the da ta requi red the application of the analytical techni que s

to both the nonsupervisory group and the group leaders as two sepa rate

and disti nc t groups.

Additionally, a commonly held practice require s at least one data

point for each estimated para meter. For this reason, certain analytical

techniques could not be applied to all the data. Tech niques where the

amount of da ta points did not sa tisfy thi s requi rement are identified in

subsequent sections.

3. 8. 1 Frequencies and Statistics

Means, sta ndard deviations , frequenc y of responses. and the

nu mber of valid da ta responses was identified for each variable, where

appropriate, using the statistical routines available in SPSS. This ana l-

ysis allows a quick check for the data ’ s distribution and the appropri a te-

ness of the responses.

3. 8. 2 Fac tor Analysi s

Each of the instruments in this questionnaire have been factor

analyzed in other studies. In particular, Rotter’s I-E scale and the

Leader Reward Behavior ins~rument have been investigated for thei r

di rnensionality. For this reason , factor analysis was applied to the

3’

_ _
 - - - - _ _

~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~ -~~~~~~~~~~~~~ -

survey da ta to test for the di mensi onality of each of these measures for

this sample.

The SBDQ could not be factor ana lyzed because the a mount of da ta

was not large enough to meet the a forementioned restriction.

There are three decisions to be made when utilizing factor anal-

ysis. The method of factor analysis, the method of axi s rotation, and the

decision rule to apply to the initial fa ctors must be identified.

The SPSS manual identifies five factor analysis methods. Princi-

pal factoring without iteration was selected as the fa ctor anal ysi s tech-

nique for this research because of its frequency of use in this type of

research and the recommendation of the writers of the SPSS manual to

those individuals who are unfa miliar with factor ana lysis methods (Me ,

et al , 1975).

The VARIMAX orthogona l method of rotation was also selected

because of its frequenc y of use and the same recommendation s by the

authors of the SPSS ma nua l.

Har ma n (1976) identifies two decision rules to apply to select

factors which ha ve any practica l significance. These decision rules are:

1) the scree rule , and 2) the rule which proscribes retaining a number

of factors equal to the numbe r of principal components whose el gen values

are greater than one. For purposes of this study , only the second deci-

sion rule was used to select signi ficant factors because of its frequency

of use.

32

b
l.
. — --~ -— —— ~~~~~~~~~~
3. 8. 3 Correlation Analysis

The next step in the analy sis process was a genera l chec k of the

linear rela tionships between the va riable s in Table 11. For those vari-
ab les approxi ma ted b y a normal distribution , the Pears on Product-

Moment correlations were calcula ted using the appropriate SPSS routine.

It was expec ted tha t the productivity variables would not app roxi-

mate normal distributions (Stevens , 1976) , therefore, nonpa rametri c

techniques within SPSS were used to calcula te variab le correlations.

For purposes of this researc h , if nonparametric correlations are re-

quired, the Spea r man correlation coe fficients were calculate d and used

for determining variables relationships.

3. 8. 4 Regression Analysis of Predicto r Va riable

Once the degree of linea r correlation between variables had been

de ter mi ned, an investigation was made on the amount of variance di s-

played by the criterion va riables. Dee to the expected nature of the

productivity distributions, regression anal ysis was used to analyze only

the job satisfaction variable.

Although a number of regression techniques are available , for-

ward stepwise inclusion was selected as the most appropriate regression

method for this ana lysis. Forward inclusion was used on the job sati s-

faction variable with the individual va riable s and supervisory va riable s

serving as the predictor variables.

33
~ ---. .
- - -._-
- . , , - - — - -.,-— -- -
~~~~~~~ ~~~~~~ ~~~~~~~~~~~~~~~

Due to the small number of da ta points available on gr oup leaders ,

regression analysis was not used to analyze the job satisfaction of these

individuals.

3. 8. 5 Test of Means

As a result of the analysis it is hoped tha t distinctions may be

made between productive and r productive indi viduals. In an a tte mp t

to determi ne differe nces within laboratory groups. tests of means were

utilized to test for diffe rences in the individual and supervisory variables

be tween the productive and nonproduc ti ve groupings.

Due to the robustness of the t test (Boneau. 1960), this test was

used to test for significant differences between means for both the pro-

ducti ve and nonproducti ve nonsupervi sorv scientist s ,. ”engineers and for

both the productive and nonproductive group leaders.

3. 8. 6 Discri mi nant Analysi s of Productivity

The fina l ana lysis technique used was disc riminant analysis. This

analysis was based on a collection of variables and was an attempt to

derive disc riminant functions which could allow the classification of the

nonsupervisory scientists/engineers into productive and nonp roductive

categories.

Although a number of reasonable SPSS techniques existed , vari -

ables which mini mi zed the Wi lk ’ s la mbda were selec ted with an F cri-

terion of 3.0 established for both inclusion and deletion of variables.

Finally, the prior probabilities used in this study we re based on group su e.

34

-.
. -,—-- ----...-—-—-—— - - - —. -_ _ - - - —_ - - - . - - —. — _ - —--—- —-
~~~~
- -
~~~~~~
_-- .. -.---.--—
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 —,.-- - - ._ .- -..- __. .
~~~ ~~~~
~~~~~~~~~~~~~~~~~~

For the reader unfa miliar with the di scri minant analysis results,

the maximum chanc e criterion was provided in the results as a guide to

the merit of the discri minant fu nction classificati on results (Mo rri son ,

1969).
Discri minant analysi s was not applied to the group leaders be-

cause of an inadequate sample si e for the technique.

~

35

~~~~~~~ - - -_ -__ -- . -- ~~_ , , ..

.
 CHAPTER 4

RESULT S

4. 1 Overview of Ana1y~is

Due to the large a mount of analysis performed on the question-

naire data, the results of the analysis will be pre sented in a manne r which

allows a logical flow from the analysi s of one area to related areas. Addi-
tionally, the a mount of the results was increased because of the need to

divide the sa mple into nonsupervisorv sc ientists /engineers and into group

leaders. Separa te results will be presented for the two groups. For

these reasons the following outline will serve as a guide to the analy si s

results presente d in this cha pter:

4. 2 Introduction
4. 2. 1 Deleted Variables
4. 2. 2 Statistical Characteristics of the Respondents

4. 3 Di mensionality of the Mea sure ment Scales

4. 3. 1 Rotter ’s I-E Scale
4. 3. 2 Leader Rewa rd Behavior Instrument

4. 4 Analysi s of Pr oductivity Variables

4. 4. 1 Productivity Statistics
4. 4. 2 Cor re lations Between O.itput Variables
- 4. 4. 3 Correlations Between Producti vi ty and Demog raphics
4.4.4 Correlat ions Between Productivity and Predictor
Variable s
4. 4. 5 Tests of Means Between Productive/Nonproductive
Categories
4. 4. 5. 1 Tests of Means Based on Demographic
Variable s
4. 4. 5. 2 Tests of Means Based on Predictor
Va riables
4. 4. 6 Discri minant Analysi s of Productivi ty Variables

36

. -- •— — • — —
I
_ _ _ _ _ _
~~~~~~~~~~~ . ~ ~~~~
 4. 5 Analysi s of Job Satisfaction
4. 5. 1 Job Satisfaction Statistics
4. 5. 2 Corre lations Between Job Satisfaction and Demographics
4. 5. 3 Correlations Between Job Satisfaction and Predictor
Variables
4. 5. 4 Regression Analysis

4. 6 Analysis of Tea m Development Questions

4. 7 Summary of Results: Hypotheses Tested

4. 7. 1 Comparison for Nonsupe rvisorv Scientists /Engineers
4. 7. 2 Comparison for Group Leaders

4. 2 Intr oduction

A prelimina ry analysis of the returned questionnaires resulted in

the deletion of one p roduc tivi ty variable from further ana lysis. In addi-

tion. the productivity variables were appro xi mated , as expec ted, more by

an exponential than a normal distribution. As a result of these distributions .

the nonparametri c Spearman corre lation coefficient was calculated for the

productivity variables.

4 2. 1 Deleted Variables

One productivity variable was deleted from further analysis due to

the questionab le validity of a nu mber of responses. The variable “oral

presenta tion to technical or professional audiences” (PVAR 7) appeared

to be interpreted differently than was originally intended by the researcher.

The actua l number of suc h presentations reported by some of the respon-

dents was inordinatel y high for a two yea r reporting (fo r exa mple. fifty

presentations). Since thi s variable was designed to measure “ ora l pre-

sentations ” at professi ona l symposiums, it was clea r tha t thi s variable

37

~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 —~~~~~ -. -
.
~~. — -~~ -_ -- _
-~~-- .— ~~~ — .- - .- - - -

was not interpre ted by the respondents as it was originally designed and

thus, deletion was valid for thi s va riable .

Thi s inflated productivity valu e could have been caused by either

the respondents reporting productivity for a time period greater than a

two year reporting period or by including any form of briefing or presen-

lation as a part of thi s output.

The former proble m was taken into account during the initial anal-

ysis of the que sti onnaire. A maxi mu m number of fi fteen published papers.

ten unpubli shed manuscripts , 20 technical report s or technical memoran-

du ms , 20 new or improved processes (PVAR 5) and 20 ha rdware/software

specifications (PVAR 6) was considered a rea sonable maximum amount

which could be produced in a two year period. Respondents which ex-

ceeded these quantity values were deleted from any ana lysis dealing with

produc tivity as a relatively continuou s measure, but were used in ana lysis

dealing with nonproductive or productive dichotomies. This deletion re-

sulted in the loss of three nonsupe rvisory scientists and engineers and one

group leaders for the correlation analysis and did not significantly affec t

the results.

4.2.2 Statistical Characteristicsof the Respondents

Tab les III and IV identify the general demographic characteristics

of the nonsupe rvisory scientists/engineers and of the group leaders sur-

veyed. The codi ng scheme for these variables may be found in Appendi x B.

38

- .- - - — - - -- ...-- .. . .- .. . .. - - - -_ . - -_
 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - ..

TAB LE III

Demographic Characteristics of the Respondents

Va riable 1 Mean /Standard Deviation/Range

Abbrevi a ti ons Nonsupervisory Group Leaders
_____________________ N=272 N — 54

AGE 36. 6 ./9. 1 ‘22-64 42. 4 ‘6. 5/31-64

YRSSEC 5. 9 “5. 2 /0-27 7. 6/6. 8/0-26
YRSIS 3. 5/4 . 0/0-20 2. 8 ‘3. 0/0-13
NCWR 32.0/37 . 2/0-100 16. 6/21. 4 ‘0-75
NCWD 55. 9/39. 5/0-100 34. 3/33. 6/0 -100
CONM 22. 2 ,’27 . 4 ‘0-100 9.9/13.0/0-50
YEXP 12.0/8 . 4/ 0 -38 17. 9/7 . 6/0-40

I See Table II for explanations of abbre via tions

39

_ _ _ _ _ _ _ _ _ _
- -,.~~~~~~~
~ — .-— - -. ,_ -_- — - . -,
~~~~~~~~ - ~ ~~~~ ~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ —_ .-- -— ~~~ .~~~~~~

~~~~~. O O C ’ ~~ c c — — 0~~~ 0~~~ ’~~~~ C C ~ J ~~~~~

\C

~~~~~~~~~~~ ~~~~~~~~~~~~~~
:
~~~

C
‘-~ Q i~7 C
— I
__I__ CL

< -~~
C
F- ‘-
~

I’)

U
—
z
.E 9’
~ ~3
‘

~~ ~~~~~~~~

.~o Q)
~~
.
0
Q Cs
~~~~ - Li ~ a) —~~~
S..
. 0-
rJD 0 0 0 0 0 ~
~~~

~~~~~~~~~~~~~ ~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~

z .

~z
~~~
F-
9

Cr.)

40

_. _ -
,_ T —. - -
... _. _ : i .. .
~~~ ~~~~~~ ~~~~~~~~
4. 3 Di mensionality of the Measure ment Scales

In order to test the di mensionality of the measure s for thi s sample,

princ ipa l factoring withou t iteration with va ri max rotation was applied to

Rotter ’s I-E scale and the Leader Reward Behavior Instru ment. The

followi ng criteria were used throughout the factor analyses unless other-

wise noted: 1) the nu mber of factors re tained was equal to the numbe r

of principa l components whose elgenvalues were grea ter than one and

2) items were considered to “load ” on a rotated fa ctor if their loadings

on that factor were 40 or greater. In each case , previous studies pro-

.

vided data which allowed the comparison of the factor ana lysi s conducted

on this sample with other resea rch efforts.

4. 3. 1 Rotter ’s 1-EScale

In his formulation of the I-E scale, Rotter conducted two factor

analyses (Rotter, 1966). The first anal ysis, invo lving some 400 cases.

indicated that much of the variance was included in a genera l fac tor.

Se vera l additional factors involved only a few items, and only a small

degree of variance for each factor could be isolated. A si mi la r result

was obtained by Franklin (1963) where all of the items loa ded si gni fi-

ca ntly on a general fac tor which accounted for 53% of the total scale

variance.

The Pearson product moment correlati on coe fficients obtained

by correlating the indi vidua l Rotte r questi ons -re containe d in Appendix C.

41

- - _- -- _ -. . - . -
~~~~~~~~~~~~~~~~~~ ~~~ ~~~~~~~~~~~~~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 ..-,—
-~~~ —— —~~ ...
~~~~~
- .- -*. -
..---- . --- .. - , - . , . --—

These coefficients revea l tha t many questions were significantly inter-

related. A factor analysis of the 23 questions which make up the to tal

Rotter score resulted in seven factors.

Appendix D contains the results of the factor analysis including

the eigenvalue for each fa ctor and the va rimax rotated factor ma tri x.

Tab le V provides a summa ry of the factor analysis. The items

(Ri to R29) correspond to the questions (i to 29) contained in Appendi x

A (Part B).

TABLE V

Rotte r I-E Scale Factor Analy sis (N=273 )

Factor I Fac tor 2 Factor 3 Factor 4

R3( . 43) R4(. 57) R2( . 67) R9(. 61)
R 12(. 73) R6( . 64) R18(. 43) R13(. 4 )
R 17(. 67) Rl 1(. 53) R2 1(. 76) R15(. 59)
R22(. 73) R16(. 71)
R29(. 46) R25(. 45)
R28(. 64)

Factor 3 Fac tor 6 Factor 7

— -

R7(. 61) Rl0(. 70) R5(. 60)

R20(. 66) R23(. 75)
R26(. 61)

The results of Table V clearly indicate that the Rotter I-E scale

is multidi mensiona l for this sample, a contradiction to Rotte r ’ s claim

that the scale measures the one dimension of genera l control . Other

42
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___

researchers (Guri n , 1969: Cherlin, 1974) have found that the characte r-

istics of the populat ion sa mpled may affect the struc tu re of the factors

obtained. In particular . Cherlin found two factors whic h he refe rred to

as general control and political control. If these sa me terms were app li ed

to the present analysis, the only clear-cut factor is Factor I which cor-

responds to Cherlin ’s political control. The remaining factors reveal no

distinct separation of items i nto categories. Although the scale does not

measure one di mension and the refore it would be invalid to iden ti fy inter-

na ls or externals, it was dec ided that the total Rotte r score might provide

a valuable prediction of product ivity and job satisfaction and, there fore.

was used in additional ana lyses.

4. 3. 2 Lea de r Reward Behavior In strument (L .RB I)

The Pearson corre lation coefficients obtained by correlating the

individual LRBI questions with each other are contained in Appendix E.

These coef ficients revea l tha t many questions were signi ficantl y inte r-

related. A factor analysis of these questions resulted in thre e new factors

being identified with ei genvalues greater than one . Very litt le variance
was exp lained by the third facto r and it was eliminated in order to compa re

the results of thi s study with the original analysis by Johnson (1973).

Table VI provides the eigenvalues and associated perc ent of explained

va riance.

43

_ _ _ _ _ _ _ _ _ -,
_—,_-.
. —.
~~~~~~ ~~
 .
—- -
_ _ - ~~~~

TABLE VI

Eige rivalu es of LRBI Factor Analysis

Percen tage Cu mulati ve

-

Factor Eigenva lue of Variation Percentage

8. 81236 40. 1 40. 1

2 2. 53789 11. 5 51. 6
-
3 1.07063 4. 9 56. 5

The da ta included in Table VII contains the rota ted fac tor matri x

derived fro m this sample and the factor ma tri x derived by Johnson (1973)

in the ori~ inaI factor anal ysi s of the instrument.

TABLE VII

Leader Reward Behavior Factor .\na lvsis

Samp le Anal ysis Johnson Anal y sis

_ _ _ _ _ _ _ _ _ _ _ _
Factor I Factor 2 Factor 1 Factor 2

LI . 7~ -.08 . 71 -.07
L2 -.2 2 .52 -.17 .49
L3 .7 1 - . 15 .70 - . 10
L4 .58 .30 .52 .32
L5 -.04 .5 0 -.06 .33
L6 . 7S .07 . 72 .06
L7 .00 . 68 - .01 .56
L8 • 7 .13 .70 .07
L9 .55 .03 .62 -.03
Ll0 .83 -.01 . 77 - .01
LII .25 . 73 .23 .60
L12 . 19 .71 .08 .65
L13 .76 .06 .75 - .02
L14 .81 .03 .75 .01
Ll5 .84 .00 . 78 - .01
Ll6 . 70 .2 2 .59 .24
L17 .20 .53 .12 .3 9
L18 .66 .38 .56 .4 1
L19 .62 . 11 .6 4 . 13
L20 .6 7 -.07 .62 -.06
L2 1 .63 .2 1 .41 . 17
L22
- —
.74 .25 .51 .23

44
_ _ _ _ _ _ _ _
_ --- . --- - - .
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~ ~~~~~~~~

As identified by Johnson , Factor I was terme d positive rewa rd

behavior and Fac tor 2, punitiv e rewa rd behavi or. The facto r analysi s

results indicate tha t L2, L5, L7, Lii , L12 , L17 load onl y on the second

fa ctor while the re ma ining items load only on the first factor.

Clearl y, the results of the factor analysis revea l two distinct

factors: Factor I which can labeled “Positive Rewa rd Behavior ”, repre-

senti ng the re la tionship between good performanc e and lea der-administered

rewa rds; and Factor 2 which can be labeled “Punitive Reward Behavior ” .

representing the relationshi p be tween low job performance and leader

correc ti ve behavior.

4. 4 Analysis of Productivi ty Variables

The anal ysis of producti vity for the respondents was based on six

ca tegories. Table II previousl y identified those six categories. The

seventh productivity va riable identified in Table II, “oral presentations” ,

was dele ted from anal ysis for reasons previously di scussed. Since the

analysis was conduc ted on two groups (nonsupervisorv scientists/engi neers

and group leaders), the analysis results for these groups is presented

separately.

4. 4. 1 Produc tivity Statistics

Tab le VEIl identifies the statistical data for the six ou tput ca te-

gories for the nonsupervisory scientists /engineers and for the group

leaders.

45

.— -— .—— -. .— ——— —— _ _4•••_.__—. _ ,__ - .

— ——— — .— —. -——
~
 .-- - .. _ . _
~~~ ~~~~~~~ -- -~~~-- -
_
-~ -

-7
_ 1-
~~ ~

‘C
Oc, ~~~
~> -.
,
~~
~
0’
r— C’ —~ •
v~ Lf~ ~~
— ~~- ~~— ~~ N -
~~ ~~~

..
S
CI C’ ‘ ‘—.
~ — ‘C
~~.
— . — -
~~
~~~ ‘
~~~ N ~~
‘
~~~
C’
~~~~~~
0 ~~
-
‘— --
~
—
.
~~~
—
~~~
-
_
-_ -
N L( —
“:
~~ ~ ~~
cc ~~ ~~
— ~

a)
Ct
2)
~
__________________________
ct:

cc
‘-I . • N c~
0 N C’ •
-~ Ct C
‘ ~~

~~ ~,

..
0
> Ct
~~~~
—. — —_ -_-
- --
C
~~~~ ~~~~ -
— C N N
~.c
~ ‘C
N C’ C
~~
~~~
C
:

S.
—
C
a)
>

a)
— c.’1 C,, ~~~‘ ‘C

< < < < < <

46
 4. 4. 2 Cor re la tions Among Output Va riables

Table IX identifies the Spearman correlations among the six ou t-

put categories for the nonsupervi sorv scientists /engineers and for the

group leaders.

TABLE IX

Correla ti on s Among Ou tputs

Nonsupervisory Scientists /Engineers (N=26 9)

PVAR 1 PVAR2 P\TAR 3 PVAR4 PVAR5 PVAR6

PVAR 1 1. 000 . 256* . 231* - . 064 . 144* . 094

PVAR 2 1. 000 166 ~k .15 1* .2 67 * . 182 *
.075 . 207 *
.

PVAR3 1. 000 1 74~

FV AR 4 1. 000 . 091 .l 3 4 ~~
PVAR5 1. 000 .26 1*
PVAR6 1. 000

Group Leaders (N=53)

PVAR 1 PVAR2 ~ PVAR3 PVAR4 PVAR5 PVAR6

PVAR 1 1. 000 394 * . 199 . 220 . 086 - 073
. 091 . 104
. .

PVAR 2 1. 000 . 237~~ 290~~

.

PVAR3 1. 000 • 379* . 4 09* 138

.

PVAR 4 1.000 .318 * .294**

p\TAJ(5 1.000 . 285**
PV AR 6 1.000

*~~ less tha n or equa l to 01

.

**~~ less than or equal to 05.

47

_ _ _ _ _ _ _ _ _ .
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 Tab le IX indicates, for the nonsupervisory scientists/engineers,

that significant correlations exist between the publication of journ a l

articles and unpublished manuscripts , the publication of journal articles

and the writing of technical reports, the publication of journa l a rticles and

new or improved processes, products , techniques, and pa tents or pa tent

application , the writing of unpublished manusc ripts and the writing of

technical memorandums, the writing of unp ublished manuscripts and new

or improved products , processes, techniques and pa tents or patent appli -

ca tion, the writing of unpublished manuscripts and the output associated

with contract monitoring , the writing of technical reports and new or

i mproved p roduc ts , processes, techniques , and patents or pa tent appli-

cation, the writing of technical reports and the ou tpu t associated with

contract moni toring, the writing of technica l memorandums and the ou t-

put associated wi th contract monitori ng, and new or i mpro ved processes,

products, techniques and patents or patent application and output associ-

ated with contract monitoring. Although these correlations are signi fi-
cant, they are weak , indicating little association among the va ri ous

produc tivity variables under considerati on. Thus , it would appear to be

iate to analyze the se variables sepa rately.

appropr

For the grou p leaders , Table IX indicate s that sign i ficant, although

weak , correlations exist between the publicati on of journal articles and

unpublished manuscri p ts , the writing of technical reports and the writing

48
of technical memorandum, and the writing of technical reports and new or

improve d processes , produc ts. techniques and pa tents or patent applica-

tion for the group leaders . Thus, as was the case with the nonsupervisorv

scientists/engineers, it would appear to be appropriate to anal y7e these

variables separa tely.

4. 4. 3 Correlations Among Produc ti vi t and Demographics

Table X identifies the Spearman correlations amon g the six output

categories and the nine de mographic variables for the nonsupervi sory

scientists/engineers and for the grou p leaders.

TABLE X

Productivity Correlations With Demographic Variables

Nonsuperv i sorv Sciend sts / Engi nee rs (N 269)

,

PVARI PVAR2 PVAR3 PVAR4 PVAR5 PVAR6

AGE . 094 . 037 . 218** -. 043 . 125* . 113*

GRADE . 196** .118 * . l83** -.010 . 169** . 126*
YRSSEC .051 . 079 .175** . 135* . 133 . 226* *
YRSIS - . 025 . 034 • 118* 1 14* . . 036 . 2l0
~~
EDCTN 306**
- l65**
. . 076 - . 012 . 097 - 006
.

NCWR . 068 . 132 ~’ 130” . 126

• . 069 019
.

NCWD -
.005 - .064 - 109* . 062
. -
. 010 . 027
CON M . 068 - . 062 . 013 . 039 -
- . 067 . 212*
YEXP 138*
. . 095 . 208 ”” '’ 012 . .147* l4 8 **
.

Group Leaders~~~-53)

AGE -.085 . 018 . 000 . 078 . 017 - 015

.

GRADE . 084 . 133 . 109 - .079 . 150 - 015

.

YRSSEC -
.090 . 102 - .088
-
- 238*
. . 066 . 062
YRSIS . 209 . 075 - . 090 - . 141 - . 085 - 059
.

EDCTN .406** .409** .271 .201 .182 -.057

NCWR .446** . 148 . 162 . 150 • 169 - 042
.

NCWD -
.129 .024 -.056 -.015 . 152 .142
CONM . 252* . 230 . 197 . 088 -.011 433**
YEXP - .133 -.043 -. 003 -. 057 . 073 - . 124
*p less tha n or equa l to 05 .

~~ less than or equal to 01 .

49

-~~~ — -- --rn-- -—- ,- -“ -_ - .- - . —, . .-

~~~~~- — - ---- ~~~~ --
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 _ _ _ _ _ _ _ _ _ _ _ _ _ .- _
_ _ _ _ _ _ _ _
~~~ —~~~~~~~~~ - — - _ - . - . -. .— -_

For the nonsupervisory scientists /engineers , Table X identifies

a significant negative correlation between the percentage of ti me spent

in development and the nu mber of technical reports wri tten (PVAR3).

Whether this information Indicates that the amount of time qr it in

development activities can inhibit thi s type of output requi res further

inves tigation. The positive correlation between the amount of time spent

in research and the number of tech nical reports written indicates tha t

technical reports are associated wi th resea rch activi ties.

For the group leaders, Table X i dentifies a significant positi ve

correlation between the amount of time spent in research and the number

of published journ a l articles. Surprisingly , a sit ni ficant positive corre l-

~
ation exists between the a mount of time spent in contract monitorin g and

the number of published journa l articles.

4. 4. 4 Corre lations Among Produc ti vi ty and Predictor Variables

Table XI lists the Spearnia n correla tions among the productivity

variable s and the predicto r variables for the nonsupervisory scientists

engineers and the grou p leaders. Due to the number of missing values,
tbe sample size correspondi ng to the predictor variables is in paren-

thesi s beside the variable abbre via tions.

50

— - , ---—--—- , —~~~~
 -~~ —~~~~~
__ -~~ -_ — - -_ - _ . - .

TABLE XI

ProductivityCorrel~itions Wi th Predictor Variables

—
Nonsupe rvisory Scientists/Engi neers
PredIcto r
Variables PV AR I PVAR2 PVAR 3 PVAR4 PVAR5 PVAR6

R OT(225) .03 1 .05 1 . 062 . 096 . 138 * -

019
.

HOP(26 9) .052 . 064 - . 036 -. 095 . 040 -. 080

SBDQC (255) -. 006 .042 -.024 -.046 . 003 -
104 *
_
.

SBDQI(255) .117 * -
104 *
. -.092 -.097 -.026 -.015
LRBP (261) -
. 020 . 060 - 080 -. 031 -
. 019 - 047
_ .l35*
. .

LRBN (260) -
008
. -. 058 -.01 3 . 061 .01 8
Group Leaders

ROT (45) — .024 .089 -.08 2 .045 .057 .107

HOP (53) .020 -.224 -.010 .026 .1 86 .202
SBDQC (48) -
159. -.086 .024 .002 .071 . 128
SBDQ I (49) - . 048 -. 179 - . 088 . 032 .04 1 . 208
LRBP (50) . 2 15
-
- . 1 26 - . 004 - . 046 . 029 . 1 04
LRBN (50) -. 031 -. 117 . 284* . 000 . 249*
I See Table II for expla na tion of abbreviations
* P less than or equa l to 05 .

P less tha n or equal to .01

For the nonsupervisory scientists, Tab le Xl identifies a significant ,

but weak , correlation tha t exists between the concept measured by the

Rotte r scale and new or improved processes, products, techniques. and

p atents or patent applicati on. The justification for this correlation would

only be specula tion since difficulty exists in deter mi ning what ac tually is

measured by Rotter ’s scale.

Addi ti ona lly, two ra ther surprising results are the si gnificant

1iega ti ve correla ti ons between i n i t i a t i n g structu re and published jou rna l

articles , and Init iating struc ture and unpubli shed manuscripts. These

correlations Indicate tha t a supervisor high in initiating str.icture is

51
 -— .-~~~~~.

associated with lower output in these two types of producti vi ty. A parallel

correla tion is the significant negative relationship between punitive leade r

reward behavior and technical memorandum. This correlation also indi-

ca tes that a supervisor hi gh in punitive reward behavior is associated

with lower technical memora ndum output.

Consideration by the supervisor, however, also tends to inhibi t

one type of output. A significant negative correlation exists between con-

sideration and the output associated with contract moni toring. Thus, a

supervisor high in consideration is associated with lower contract moni-

toring output.

For the group leaders, Table XI identifies only two si gni ficant cor-

re lations, both i nvolving the ind i vi dual’ s perception of his supervisor’ s

negative reward behavior. Nega ti ve leader rewa rd behavior is signifi-

cantly positively correlated with both technical reports (PVAR3) and

new or improved processes, products , tech niques, and pa tents or pa tent

application (PVAR5). These positive corre la tions indicate that the per-

ception by the group leaders tha t thei r supe rvisors are likely to use puni-

tive reward behavior for low per f ormanc e in these two categories tend s

to be associated with these two types of output.

4. 4. 5 Tests of Means Be tween Producti ve / Nonproductive Ca tegories

Although the correlation coefficients identified many significant

relationships between the productivity variables and both the demographic

and the predictor variables, no Consistent set of variable s were fou nd

52

_ _ _ _ _
_ _
_ _ _ __ _ _ _ _
~ --~~~~~~~~~~~~~~~~~~~~
- ~~ -~~~~~~~~~~~~~~~~~~~~~ - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

which were uniform across the enti re set of productivity va riables. Due

to the exponential distributions associated with the prod uc tivi ty variables

and the number of responses with no output, the two groups (the nonsupe r-

visory scientists /engineers and the grou p leaders) were divided into pro-

ducti ve and nonproductive groups as previously defined. Tests of means

were then used to ana lyze group di fferences. Due to the volume of anal-

ysis material, the results are presented in two sections: first , producti ve !

nonprod uctive versus the demogra phi c variables, and then , productive!

nonproductive versus the predic tor variables.

4.4. 5. 1 Tests of Means Based on De mographic Variable

Table XII identifies the results of tests of means between the p ro-

ducti ve and nonprod uctive categories based on the de mographic variables

for the non supervisory scientists /engineers.

Tab le XII identifies GRADE as being the demographic va riable for

which a significant diffe rence exists between the productive and nonproduc-

tive categories for each produc tivi ty variable except for technical memo-

ra ndums (PVAR4 ). Additionall y, a significant di fference exists between

the productive/nonproductive groups for published journa l articles (PVAR 1)

on EDCTN and YEXP, for unpub li shed manuscripts on EDCTN, for tec hni-

cal reports on AGE , YRSSEC , and YEXP , for new or improved processes,

products, techniques and patents or patent application on YEXP, and for

ou tput associated with contract monitoring on YEXP. No significant di f-

ferences ex
isted between the productive and nonproductive groups for

technical memorandum (PVAR4).

53

.~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~
 — -
- - ..-

if ~~ tr:~ — .c
tf ~ C C
~~~
N ~ ~—
~ c
—

C C tr
~ -. .
- .-. —
.
~~-
s
—
C’~ N C’~ 0 tf N C
— ‘.
~ ~~~
‘
~C
~ C
>
~
_z
— _ f t

— N

— N C’ ~
~~
fl
N
C
—
C C ~~~
.
c—i
—
~~~

.
~~~
C
~~~
< ~~~~,
— -— -- — .-- >
~~ ~ — — ~
c--i C C if) ‘.0 ~~ c-- C’ \C C tf : N
—

~
>
- -€
.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-

~~

——
‘.Q
_
0 ~~ N C’
C’
~~
C~~~ C~~~ _ N . O \0 C\ O Z Z
>‘ it) ‘.0 C’) c~) c--i c—i — c - - i >
N ‘-0 ~~~‘
c’~~c-~ ~~ ~~
~~~

~~~~

• - -- 2 <
~ ~~
-- .
— ~ .~~~ c-i — C’ c-I c-~ c-i ~~ ‘.c N ~~~ — .

R ~~~~~~~~~~~~~~~~~~~~

--. -- .— ---. --.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
—
. ~. - ~ ~ ~
If) N C’ C’-1 N c--i it) If) C

——
. ‘~~~
‘.~~
o
~~~
. .
> ‘.0 ‘.
0 Cl)
Cl) Cl) ~~ If) ‘.0 Cf) Cl) C~’1~~ — ~~
‘
C’, C’. c~
- Cl)

0’
~~
Z —c. ’
~~
< ‘— — II~
_

B o c
c--I ~~~~

>,‘~~~2
~~~

——
~~
> Cl) L:.
~~~ ~~~~~~~ <
C’, CI)

——— —
~- 0>
ij c;~ n~
~
_ _ _ _ _ _ _ _ _ _

>

—
*
C C
c-i c—i c—i c--I — c--i c--i
~~
14 ~~

g
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

z
C) ‘

F-
.
Q
~~~ Cl)
(I) ~~
—
Cl) Q X o
~~ ~
0
>. < 0 ~‘- ~
-
~1
N—

54

—- -• -
-- - —-- = = - . —
.-

~ ~ ~~
- -
~~~~~~~ ~~ ~~~
 Tab le XII identifies the results of tests of means between the ,ro-
~
ducti ve and nonproductive categories based on the de mographic variable s

for the group leaders.

Tab le XIII identifies significant diffe rences between producti ve and

nonproductive group leaders on the various producti vi va riables. EDCTN

~~
was found to be a signi ficant variable for identi fying differences between

produc ti ve and nonproductive group leaders on publi shed journa l articles.

unpubli shed manuscripts, and technical reports. No si gnificant differenc e

~
were identified between productive and nonp roductive group leaders on tech-

nical memorandu ms and new or i mproved processes, p roduct . tec hniques

~
and patents or pa ten t application. YEXP was found to be significantly dif-

ferent between productive and nonproduc ti ve grou p leaders on the output

associated with contract monitoring.

4. 4. 5. 2 Tests of Means Based on Predictor Variables

Table XIV identifies the results of tests of means between the p ro-

ducti ve and nonprod uctive categories based on the predicto r variables of

ROT , i-LOP, SBDQC, SBDQI, LRBP , and LRBN for the nonsupe r vi so: scien-

ti sts and engi neers.

55

______________ _ _ _ _ _ _
~ .-
.- . - - - - -, . — - - --- -- . . . - ---
,-
- - . - -
- - ~ ~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~ ~~~ ~~~

(V--i Cl) N If)

c-i If) ~~~

C’ Cl) Cl)
~~ — c- Cl) If) C
C —
. —
.
- —
. —-
. —-
—
. -.
.
-
-z
~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~

>
‘— -r -i~
— ‘.c 4
~~ c-~~~~~~
~ ~
—
z I’

1’
C’ c-i .‘o —
if) cc ~~ . c-i
N-
~~
Cl) ~~
N c—I If) > >
— —: - —
< <
ccC ’ f) 0 1 t) Lf) 0
< ~~~~~~~0 -~

_ _ _ _ _ _ _ _ _ _ _ _ _ _

~~
I—, ~~~
~~~~~ <
~~~~ — — -.-

~~~~~~~~~~~~~~~~~~~~~~~~~~~~
‘;_
—
-
— >-
~
>

‘
~~~
Cl) .~ if) If) C’ it) C1) c--i c-i Cl) cc N
. -

~~~ -
c a
-
~~~ ~~~~~
c ~~
c-i Cl) cc ‘.0 N- 0 U
~
U~
Cl)
C’ N
Cl)
If)
it)
Cl)
C
it)
C’
- ‘Z
—‘ —v z -
~
~
~~~ —c —: ~~~~
—
~~~ ~~ ‘ - i~~~
’
c--I .’0
— Cl) c-i it) C’ -i ‘.C cc
= :1- —
- - —
~~~~ ~~
~ ~~~

~~ ‘1
— ~

~~ cc
— cc ~~ ‘-C
Z - .
~
‘-
~~~
-
~~~ ~~~~ ~~~~
$
c N N ’ .~~ N~~~~~~~~~~~~~~~~
~
1 1
~~

.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

I c—i
~
Cl) ~-
<
l-~

— c--i Cl) Cl) ‘.~~ c—i ‘r ‘C N cc i f) ”

c- c-

— —
-
~~~~ -

0
c-i — c-i c-i — — c-i
~~ ~~~~~~~~~
c-i — c-i ~
C
z —
~~~

-
C)
Z
F- ~.-
~ ~~ L~ >< ~
~
c~ 0 0
> < ~0 > ~~ C~~ >~ ~

56

.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - . . ~~- . -~~~ _ _ _ _ _ _ _ ~~~~~~~~~~~~~ ... ~~~~~~~~~~~~~~~~~~~
 — - - -- —
- -~~~~~~~ -~~~~~~~~~~~~~~~~~~~ -- -- -~~~~ --- ~ - -- - -- — - - -- --
~~~~
~

If) Cl) If) C’ C.-’ c-i

C’ If)
C’
it)
‘C if) N ~~
~~ ~~ C’ Cl)
~~ . • ‘

— — C’
N

—————
- ~~
- -
~~~ ~ - .

< C’.
cc N . 0 _< < N ’C N c- 0
- If) • Cl) ~~~ ~~~~
c—i •~~
• c-i - ‘0
0
• c-i
> . - — ~. ‘C c-i
-“- ~~ c~~~c --~. cN --
— cc N — C~ ’ ~~ c’.
~ cc -
c-i
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ cc~~~ .~~~~~~ N
~~~

Lf) 0 Cl) ~~ C’
If) ~~cc~~~~N
_
_
cc ‘C ~~~
-
cc N ‘.0 ‘0
~~ N- cc N ~~-c N
—
‘C ~~ ‘
~~~
— ~~
C’ — C’ — ~~
C’ — C’ N — C’ ~~ cc cc

— _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

C’ cc 0 ~~~
Cl) c--4 If) ~~

(N c- • l) C

>< C
~~
-
..
¼-~~, .‘0

‘C —
-- c c -~~~
(N ~~~ - cc
—~~ N-
C’
cc ~~ ~— . C’
—.
—
---Z Cl) — c c — . C’-i ’—
C’
—N—
‘ -

l~— C’ z ~ C
‘

—
r ~~~ - CI~ ~ ‘~~-
-- . C’ ~ -- ~~-- —~~N ~~~
.,~. ~~— C)l -
— ~
- -- c c —.- c-i
— cc — - --
— .— ‘
Cl) Cl) C’ c-i N C’- ~— - (N
~~
-i C
~ ‘Cl
.

—
r- ’ C — cc N- cc ~~ cc Cl) c-i c- c Cl) c-i Cl) C—1 Cl) C’i
~ )
-ç
—
Z T
cc
< > C) ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _
—
_ _
_ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

-
~~ C
Cl)
Z C’
C’
If) ~~~4 ~~ If) -‘C c-i
cc ~~
.-~~~~~~~~~~~
~~ C N- If)
‘ 0 N - N cc
‘ c
— — . _•5_ __ _
—5
---
_•___
5
-5-
--
--5
—
~~~~~~~~~~
. —-5
—-
__ _
5 - --5
- —-5
C
-

~~
) N-~~~~~O C’~~~~~’C O cc ’0 c-~~~~ C~~~~~~~~~~~~~~~~ ~~~~cc
> > - ~~~~~ -.
‘‘ -c-i
cl) •~~~~~~~~~~~~~ C c - _ I

cc — cc —
~~~~

— cc — — — — — — — — — — —
~ ~~~~ ‘
LI) 5-5- C’ , If) If) - c c ~~~~~ N ~~ ‘- c-i If) if) N ‘f). C’•) . -. . _ c-i
—
Cl)
cc ~~~ ~ ~
~~~ N ~~ —
C’ N-~~~~ ’C ’0 c-_ I C ~~
— N — cc N- cc N-
~~~~ —

~~~~~~~~~~~~~~~~~~~~~~~~~

— (N — (N — (N — (N — (N — (N
— (N — (N
— (N — (N
— (N — (N C

— c—i
C) 0.

57

L .
~~~~~~~
_ _ _ _
. - -
~~~~~~
.-
~~~~~~~~~~ -.-
_ -— - .-~~~~~~~-— - S~~~~~~~~~~~~~~~ . - - - - 5~~~~~~ - . -~~~~~~~~~~ - - — - --~~~~ ._ _ - -

Surprisingl y , onl y one diffe rence was found between the productive

and nonproduc ti ve groups over all predic tor va riables and producti vi ty

variables. ROT was identified as a distinguishing va riable for new or

improved processes, products, techniques and pa tents or patent applica-

tion (PVAR5). A sign ificant di ffe rence was found between the p roducti ve
or more external group as defined by the f-E sca le and the nonproductive

or more interna l group as defined by the I-E scale.

Tab le XV identi fie s the results of tests of means between the pro-

ductive and nonproductive categories based on the predicto r variables of

ROT , HOP , SBDQC, SB DQI, LRBP , and LRBN for the group leaders.

--- 5 - -
~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 -, - --- - .----- .--—.-- -— - .. --- , -

~~~~~~

If) Cl) It) C’

‘.
0
‘.0
)
‘ ‘-C N ‘ — r
‘.0
s c-i
o ~—
)4 N-
~~~
C ~~ -
~~ ~. c- C’
—: ~ N
C’ ~~~

.:
_ ~~~
— • - -
~~~~~~~ ——.~
~~~ ~ ~ ~~ ~~~~~~~~~ -5-~
-•-
<
•
(N ~
~~ N~~ O C~~~~~~~ ~~~~~~~~~~~~~~~~~~ C’.c c-i cc ’C
C’ C C N It’ 0 cc C’ ’C ~~ ‘.c N- ’C -
‘ ‘ 0~~~~ ~~ N - NC ’ C
) l’ cc C’
N- c-~ (N ~~
0 _
~~~‘ ‘C C~~ C’, C’~ N - c c c—i
(N

— — — — — —
“‘C ’ cc — — cc — C’ c- C’ c-~ C’- . c--i If) ‘~~~ LI) It) ‘. It)
(N — Cl) ( N C l) Cl) Cl) — Cl) ~~
c~) .’—
C ~~
Cl) — Cl) Cl) ‘C C’) ~~~
~~ -. C’)

Cl’ — Cl) C C’
- ‘

0.
,
~ ~~ cc Cl) C’ 0 If) C’ N
~~~
if)

5- .
•

~~~< ‘ 5
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
(:_) _ • cc C . . . • . iI) .
~.,
C ‘.~~~ (N — N’ C Cl) cc
C ’c c • C C’~~~~’CCl)~~~~~~~
,.Z C’,
~~~
‘ N- cé
~~ c-~ c-i ~ (N
—
N-

~~~~~~~~
-
~~ C
—
~
,

~~~
~p
~~ c--i (N C C’ — C — C —
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I) _ — a,-, h
~~~~~~~ - cc cc cc
J (N,..C’ ,C - C’ _ . c c ~~ cl)._ c cl) —
C’
~ Q l ( N ( N ( N ( N C ’ ) ( N C(N ’)
- U) ~~~~~~~ ~
~~
cc
c~. ~~~~~~ _ _ _ _ _ _ _ _ _ _
- _
C C
C
z ~
.
5-—
~ — c-i cc C C if)
C) - —.
~~~‘ cc ~~
— if) — C C’ cc
E ’ ~~
_
~~
> c--I 1!) C’- cc
~~ C ‘0

.-
~~~ ~~~
LI • - ~~
-~~ ~~ ~ — —
~~~
_
~~~~
—_ — —_
- —_

’ ’N C Lt)
(N C l )c c <5- N- Nc ~~
c_
~~
‘C
~~~~~~~~~~~~~~~
>. “ cc •
C ’c c
~~~~~ ~
ç
~~~~~
> ~~~~~~
~~~~
. . C’
•
. . . ..’~r C . ’
~~~~ c-~ c-~~ ~~~~
- c ~~ C’ — N cc Cl) cc N- (N (N — C—i N Cl) Cl) cc N- (N (N
‘.

C’ — Cl) C C C C’. (N C’ (N r) N N . If) II

) ’.0 If) i f)
Cl) (N cc C’ Cl) (N (N (N c- (N (N —
If)
c--i — (N (N (N (N ~~(N (N (N (N (N ’0 C’i (N

*0. L I0
C 0.
0 — — — (N — (
N — (N — (N — (N — (N — (N (N — (N — (N
—
~
0

— (N
C) C.
—
,
~~
c—’ c- ~’
,
_
Z .
—
C— ’, Q-. Z
o
~~~

~~~~~
-‘ ‘-LI cc LI ‘LI
Cl) -~ CI) Cr)
~~ ~~

59

~~~~~~~~~

-. -5- .— . _ - -- —-- . , ———~~~~

 Tab le XV identified significant differences for thre e of the pro—

ducti vity variables based on the predictor variables. A si gni ficant

difference exists between the productive and nonproductive group leaders

based on the LRBP variable for pub li shed jou rnal articles (PVAR I). The

productive group leaders perceived their supervisors as exhi bi ting lower

positive reward behavi or than did the nonproducti ve group lea ders. Thi s

result may indicate a perception by the p roduc ti ve group leaders that

they are not being rewarded for performance in this category.

A significant di fference based on SBDQI was identified for unpub-

lished manuscripts (PV AR2). In thi s case the producti ve group leaders
indicate d a lower perception of initiating struc ture than did the nonpro —

duc ti ve group leaders.

For technical reports (PVAR3), a significant di fference existed

between producti ve and nonproductive group leaders ba sed on LRBN. The

produc ti ve grou p leaders perceive their supervisors as exhibiting higher

puniti ve rewa rd behavior tha n did the nonproductive group lea ders. This

result may indicate a perception by the productive group leaders that

unless thei r performance in PVAR3 is sufficient , their supervisors wi ll

rake punitive reward action for low performance.

4. 4. 6 Discri minant Analysi s of Produc ti vi ty Variables

Discri minant analys i s was applied to each of the productivity

variables in an a tte mpt to derive sets of predictor variables which

would be valuable in significantly predicti ng whethe r an individual is

60

L ... — - -
—
 - -_
- 5
- - - --— - _ --- -5 -- , -
—- ‘5
_ _ _

prod ucti ve or nonproducti ve. Additionally, di scri mi nant ana lysis was

applied to the respondents in an attempt to classify them as producti ve

or nonproductive based on these functions.

Three measures of “goodness” are presented in the analysis.

First , the significance of each variable in the discri niinant fu nc tion is in-

cluded in the app ropriate table. Second , a compa rison is made between

the percent correctly classified by the discriminant and the percentage of

respondents which could be correc tly classified by chance (ma xi mu m

chance criterion). Third , the canonical corre la tion associated wi th each

discri mina nt function is provi ded. Thi s corre lation , when squa red. is a

rough measure of the percentage of the variance explained in each of the

productivity variables by the discri minant function. In a genera l sense,

the deri ved va riables were weak, although significant , predictors of pro-

ducti vi ty and of a respondent ’ s classification.

Addi tionall y, the reader is cautioned that upward bias may have

resulted in the classification process since the sa me individuals used to

calculate the discri minant func tions were also used in thi s process. Thi s

procedure was necessa ry due to an inadequa te sa mp le sizi ng for splitting

the sa mple into two groups.

As was previously sta ted , discriminant analysis was applied only

to the nonsupervisory scientists /engineers resulting in a sample si ze of

214.

61
 - -
--
-- — -~~~~~~~~~~~~~~~~~~~~~~ --- - -- ~~~~— - -- -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

The di scrimi nant ana lysis results for published journal articles

(PVAR 1) are contained in Tab le XVI.

TABLE XVI

Discri minant Analysis of PVARI

Step
—
Signi ficant Stan dardi7ed~~
Variab le Entered Significance Coefficient

EDCTN 1 . 000 . 68129

GRADE 2 . 000 . 59628
YRSIS 3 - 000 - 45475
.

These three significant variables resulted in the correct classifi —

cation of 148 or 69. 2 percent of the respondents. This classi fication

procedure was slightly bette r than the ma xi mum chance crite rion of 67

percent , although the canonical corre lation for the discriminant function

was 37116.
- Therefore, the discri minant function is significant , although

weak in predicting this output.

The di scriminant analysis results for unpublished manuscripts

(PVAR2) are contained in Table XVIII.

TABLE XVII

Dlsc ri minant Analysis of PVAR2

Si gni ficant Step Standar i7ed

Variable Entered Significance -
Coefficient -

E DCTN 1 00l 1. 20495

62

L - ... .
~~~~~~~~~~~~ .
 - _ _ _ _ _ _ _

The one significant va riable of EDCTN resulted in an ext remely

weak canonical correlation of 02968 .

- The di scri minating fu nc tion re-

sulted in the correct classification of 150 or 70. 1 percent of the respon -

dents . Since the maximum cha nce criterion for this productivity variable

was 70 percent , the 70. 1 percent classification result is not too impressive.

Table XVIII identifies the discri mi nant analysis results for tec h-

nical reports (PVAR3).

TAB LE XVIII

Discri minant Analysis of PV AR 3

Si gni ficant Ste p Standar dized

Variab le Entered Significance
-
Coefficient

AGE 1 .000 8.78359

The use of the AGE variable in the disc ri mination function resulted

in the correc t classification of 60. 3 percent or 129 of the respondents.

This percentage was considerably less than the ma xi mu m cha nce criterion

of 64 percent. The canonical correlation for t he function is also unim-

pressive with a value o f . 24906. These results indicate that strictly by

chance individuals could be classified bette r than by use of the AGE vari-

able in a discriminatingfunc tion. Therefore, the di scri minant function

would be of little va lue as a predictor.

The di scri mina nt analysis results for technical me morandum are

contained in Table XIX.

63

-
_ _ _ _ _
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ . ~~~~~~~~~~~~~~~~~
 TAB LE XIX

Discri minant Analysis of PV AR4

I
—

Si gni ficant Step SrandardizecF

Variab le Entered Signi ficance Coefficient

LRBN 1 .028 . 79148

YRSSEC 2 .014 - - 6668 6

The classification of respondents based on the variables of LRBN

and YRSSEC resulted in the correc t classification of 56. 5 percent or 121

of the respondents. Although this correct classification result was slightly

higher than the maximum chanc e criterion of 53 percent , the canonical

correla tion was only 19937.

. This result indicates that the func tion would

aid in classifying indi viduals, but would not be a powerful aid in explaining

the percentage of variance associated with PVAR4.

Tab le XX identifies the results for new or improved processes,

products, techniques and patents or patent application (PVAR5).

-
TAB LE XX

Discri minant Analysis of PVARS

Si gnificant Step Standar dizecF

Var iab le Entered Si gnificance -
Coefficient

GRADE 1 . 009 - 81473

RGr 2 . 003 . 66264

64
 1
--

The di scriminating function based upon the variables of GRADE

and ROT resulted in a canonical correlation of 23397. The classifica-

.

tion process based on this function resulted in the correct classification

of 85. 0 percent or 182 of the responden ts , a result equa l to the maximum

chanc e criterion of 85. 0. Therefo re , the discr iminant function would not

be a valuab le aid in either classifying respondents or in explaining vari-

ance in PVAR5.

The di scri minant analysis results for the output associated with

contrac t moni toring (PVAR6) are contained in Table XXI .

TABLE XXI

Discri minant Analysis of PVAR 6

Si gni ficant
-— —

Step
—

Standardi. 7ed
Variable - -
Entered Significance Coefficient
— -

GRADE 1 . 000 1. 15417

The use of GRADE in a discriminatingfunction resulted in the cor-

rect classification of 74. 8 p ercent of the respondents. Although the canon-

ical correlati on was 03841 , the function correctly classified respondents
-

sligh tly higher than the maximum change criterion of 70 percent. Thus,

the disc riminant func ti on exp lained little of the variance associated with

PVAR 6, but could aid in the classification of respondents.

65

- - - - - -- - -
•. . ~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~
-
4. 5 Ana lysis of Job Sa tisfaction

As a paralle l to the study of prod uctivi ty , the study of job sati s-

fac tion investigates work perfor ma nce from the viewpoint of a worker ’ s

fulfillment or sa tisfaction with the j ob. This satisfaction with the job was

identified in Section 4. 4 by the tota l Hoppoc k score (HOP) as not bein g

associated with productivity either in a beneficial or detri mental manner.

Addi tionall y, tests of means revea led no sign i ficant diffe rences between

producti ve and nonproductive respondents.

The pu rpose of this section , therefore , is to ide nti fy variable s

which have a significant association with job satisfaction for thi s sample.

As in the productivity analy sis, results for the nonsupervi sorv scientists ,-’

engineers and the group leaders are presented separately.

4. 5. 1 Jç Satisfaction Sta tistics

~
Table XXII identifies the statistical data for the job satisfaction

variable (HOP ) for the nonsupe rvisory scientists /engineers and the group

leaders.

TAB LE XXII

- Job SatisfactionStatistics

Variable Mean/ Sran~ir~ Deviation/Maximum Job Satisfaction

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Nonsupervisory (N=272) Group Leaders (N=54)

HOP 18.8/3.4/26 19.8 /3. 1 /25

A comparison of the job satisfaction means indicates tha t the group

leaders are more satisfied as a group than the nonsupe rvisory scientists 1
,

engi neers. This di fference in jo b satisfaction is statistically significant at

the 05 level.
.

66
~~~~~~~~~~~~~~~~~~~~~~ ~ ~~~~
_ _ —.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~ —- -
 . - - -- - - - -
- - - --
-- 5 -_ ~~— - - - .
- - -~~~~ --- -
— - 5. - - --- - ~~~

4.5.2 Correlations Between Job Satisfac tion and Demogra phics

Tab le XXIII identifies the Pearson correlations between job

sa tisfaction and the nine demographic variables for the nonsupervi sory

scientists/engi neers and for the group leaders.

TABLE XXIII

Job Satisfaction Correlations Wi th Demographic Variables

Nonsuperv-isorv (N ~ 272) Group Leaders (N 54)
Va riables I-tOP HOP

AGE - .0 11 . 038
GRADE -.00 1 -.055
YRSSEC -. 026 .134
YRSIS - . 048 - .124
EDCTN l034’~’
.
-. 276*
NCWR 024
-. -. 003
NCWD . l425~ * .0 1 6
CONM . 022
- - 201
. 030 - 045
.

YEXP —
- .

*P less than or equa l to OS -

** ~~ less than or equa l to .01

Table XXIII identifies a signi ficant posi tive correlation between job

satisfaction and both educationa l level (E DCTN) and the a mount of time

spent in develop ment activities (NCWD). Thus, both EDCTN and NCWD

are associated with higher job satisfaction.

For the group leaders, Tab le XXIII identifies a signi ficant nega tive

corre la ti on between job sat

isfac t
ion a rid educational level (EDCTN). This

corre lation indicate s that the higher the educational level, the lower the

job satisfaction. This result is in direct conflic t with the nonsuper visory

67

_ _ _ _ _
___________________
_
— - _
~~ -~~~~~

scientists/~ ngineers where higher educational level was associated with

hi gher job satisfaction. The i mplications of this finding are discussed in

Chapter 5.

4. 5. 3 Correla tions Be tween Job Satisfaction and Predic tor Variables

Tab le XXIV lists the Pearson correlations between job satisfaction

and the predictor variable s for the nonsupervisory scientists /engineers

and for the group leaders. Due to the nu mber of missing values, the

samp le size corresponding to the predictor variables is in pa renthesis

beside the variable abbreviati on .

TABLE XXIV

Job Satisfaction Correlations With Predictor Va riable s

Nonsupervisorv Scientists ‘Engineers

Predictor 1
Variables HOP
ROT (227 ) -
. 237**
SBDQC (258) . 216**
SBDQI (258) . 084
LRBP (264) .172 **
LRBN (263) - 070

Group Leaders

- ROT (46) - 142

.

SBDQC (4 9) . 276*
SBDQI (50) -
. 031
LRBP (51) 304*
LRBN (51) . 1 26
I See Table II for Explanation
~~ less tha n or eq ua l to ~05
** ~~ less than or eq ua l to .01

68

— - _ _ — --_ - --. —
~— — i T _ .
~~~~ ~~ ~~~~~~ . — —- ---- ~~~~
_ _
 For the nonsupervisory scientists/engineers, Table XXIV identifies as

a signi ficant negative correlation between the concept measured by the

Rotter scale and job sa ti sfaction. This correlation indicate s tha t an

external as defined by Rotter would have lowe r job satisfaction tha n an

inter nal. Also , a sign ificant positive correlation exists between job

satisfaction and consideration (SBDQC). Thi s correlation indicates that

the higher the supervision is perceived as being considera te the higher

the job sati sfacti on. Finally, Table XXIV identifies a sign ificant positive

relati onshi p between job satisfaction and positive leader rewa rd behavior.

This corre la ti on indica tes that a supervisor who is percei ved to exhibit

positive rewa rd behavior for good performa nce tends to ha ve higher

sa tisfied subordi nates.

Si milarly for the group leaders, Tab le XXIV identifies significant

positive correlation s between j ob satisfaction and both considera tion

(SBDQC) and po sitive leade r reward behavior (LR BP). These positi ve

relationships indicate that supe rvisors who are perceived to be consider-

ate and that supervisors who are perceived as rewarding good perform-

ance would both tend to have highe r job satisfied group leaders.

4. 5. 4 Regression Analysis

Table XXV is a summary of the regression ana lysi s of job satis-

faction for the nonsupe rvisorv scientists /engineers. Due to the nu mber

of missing va l ues, Iistwise deletion of these values was used, thus. the

correlation coefficients necessa ry to conduct this analysis are slightl y

L .~~~~~~~~~~~~:
~
— --— - - - —
- -
~~
—
- - .- - — -----.- - -—--- ..- ---—-

di fferent than those presented in Sections-’4 . 5. 2 and 4. 5. 3. These cor-

relations may be found in Appendix F. -\s pre viously sta ted in Chapter

3. regression analysis of job satisfaction was conduc ted on only the

non.3upe rvisory scientists/engineers (N 214).

TABLE XXV

_ _ _ _ _
Regression Ana lysis For Job Satisfaction
R Sign of Beta Final
_ _ _ _ _ _ _ _ _ _ _ _ _

~~p Variable Mu ltiple Coefficient Partial F Significance

1 SBDQC . 2838 7 + 18. 58131 . 000
2 ROT . 35677 -
11.29018 . 001
—

1 See Ta5le II for ex ~ Ianation of abbreviations

Regression analysis revealed two va riable s which were significant

in predicting job satisfaction , SBDQC and ROT . [The concept measured by

the I-E scale was found to be ne ga ti vely related to job satisfaction as was

indicate d by the correlation coefficient of -

. 239 (Appendi x F). Consider-
ation by the supervisor (SBDQC ) was fou nd to be positively related to job

satisfaction by the regression and correlation analysis (r . 233). The

combination of these two variables resulted in an explanation of 13 percent

of the variance in job sati sfaction. The result indicate s weak prediction

of job satisfaction by the demographic and predic tor va riables.

4. 6 Ana lysis of Tea m Development Questions

Two questions req uested by laborato ry management to de termi ne

the effectiveness of their current tea m develop ment progra m are included

in Appendix A, Part E. These questions were designed to determine the

70

-
..i . ~~~~~.
..
 ‘1

awareness of this p rogra m by non-participants and diffe rence s between

individuals identified as participant s in the progra m versus non-pa rtici-

pants . For purposes of this ana lysis, the group leaders and nonsupe r-

vi sory scientists/engineers were combined into one sample.

Of the 326 respondents, 68 or 21 percent were awa re of the tea m

development progra m. Table XXVI identifies the de mographic and pre-

dictor variables which represent significant differences between aware

and not aware groups.

TABLE XXVI

Analysi s of Awa reness Ques don

Va riable G roup ~ Nu mber o~ Cases jrMea ns Probabil i tv

AGE 1 253 ‘37 . 2 — ’. 14L ~

2 68 ’38. Q
GRADE 1 258 ‘4. 2 ‘. 004
2 68 14.6
YRSSEC I 258 ‘6. 2!. 878
2 68 ’6. 3
YRSIS 1 258 3. 6/. 008
2 68 /2. 4
EDCTN 1 258 2 7/ . 339
2 68 ’2. 9
YEXP 1 258 ‘12. 8 1 542
2 68 13.6
I~VAR l 1 258 /. 33’. 29-3
2 68 /40
PVAR2 1 258 /. 3 1’. 259
2 68/ .38
PV AR 3 1 258 36 ’. 539
-‘
.

2 68 1.40
.

PVAR4 1 258 7.50/.095

2 68’.38
PVAR5 I 258 1~ 14’. 123
2 68’. 2 2

71

- - -_ -. _ --- ---
~~~~~~~~~~~~~~~~~~~
 —-
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

TABLE XXVI
(Conti nued)

Variable Group * Number of Ca ses/Means /Probab ilitv

ROT 1 215/7. 97. 806

2 58 /8. 1
HOP 1 258 / 18.87. 027
2 68 /19. 8
SBDC Q 1 244/71. 4 7. 118
2 63 /75. 0
SBDQI 1 244 ,/35. 9/. 207
2 64/37. 5
LRBP 1 249/80. 6/ . 015
2 66/86. 8
LRBN 1 248 /22. 77. 169
2 66 /23. 8
* Group I
~J~ot Awa re
2 Aware

Of the 68 respondents aware of the progra m , 26 or S percent of

the 326 respondents participa ted in the prog ram. Table XXVII identifies

the demographic and predictor variab les which represent significant

di ffe rences be tween pa rticipants and nonparti c ipa nts.

TABLE XXVII

Analysis of Participation Question

Va riable Group * Number of Cases /Mea r~ ‘Probabili ty

AGE 1 300 -‘37 . 6 ’. 964

2 26/37.5
GRADE 1 300 ‘4. 3,-’. 199
2 26/4. 6
YRSSEC 1 300 /6. 3’ . 264
2 26/5 . 0
YRSIS 1 300 /3.57. 000
2 26 ’l. ô
EDCTN 1 300 ‘2. 7/ . 017
2 26/3.3
YEXP 1 300 -’ 13.0 ’. 90l
2 26 ’12. 8

72
 ~~~~~~~~~ -- - - --~~

TABLE XXVI I
(Continued)

Variable Group * Numbe r of Cases/Mean/Probabili ty

PVAR 1 1 300 / .337. 188

2 26/ . 46
PVAR 2 1 300 /. 32/ . 501
2 26/ .38
PVAR3 1 300 . ’. 37!. 528
2 26/. 3l
PVAR4 1 300, ’. 48/ . 180
2 26 . / .35
PVAR5 1 300 . ’. 15 /. 100
2 26/ .31
P\T AR 6 1 300 / . 67 1. 791
2 26/. 69
ROT 1 249/8 . 0/. 338
2 24 /7. 0
HOP 1 300 / 18.8 -‘. 004
2 26 ’20. 8
SBDQC 1 281 7 I . 8 ’ . 212
2 26/75~ 9
SBDQ I 1 282 - 36 .2 1.4 7 8
2 26 ’37. 4
LRBP 1 289/81.3 7 . 088
2 26 /87. 9
LRBN I - 288 /22.8 7 .4 17
2 26 /23. 8
* Group I Nonparticipant s
2 Participants
—
Although Tables XXVI and X XVII identify some significant demo-

graphic differences between the groups, by far the most significant di ffe r-

ence between the aware /not awa re and the participants /nonparticipants

groups is in job satisfacti on. Both the aware and the participant groups

indicate signi ficantly higher job satisfacti on . Although it is i mpossible

to determine whethe r the team development program is responsible for

this di ffe re nce, the dichotomized groups vary onl y sli ghtly, but not
 _--

significantly on all other variables. No differences existed between groups

based on productivi ty for all producti vi ty measures.

4. 7 Summa ry of Results: Hypothe ses Tested

The pu rpose of this section is to compa re the productivity and job

sa tisfac tion results with the hypotheses presented in Chapte r 1. Due to

the separate analysis conduc ted on the nonsupe rvisory scientist/engineers

and on the group lea ders the comparison of results to hypothe ses will be

presented separa tely for these two groups.

4. 7. 1 Comparison for Nonsupervisorv Scientists ‘Engi neers

The fol lowing hypot hese s with appli cable results are provided:

Hypothesis 1. A positive re lationship exists between p roducti-

vi tv and job satisfaction.

Result s: This h~ pot hesis was not support ed by the results at the

01 leve l for any productivity variables (Table XI) . Additionally, no

Si gnificant di fference was determined between productive and nonpro-

ductive personnel based on job s a t i s f a c t i o n (Table Xrv) .

The results are similar to those report ed by Vroom

(1964 ) where vi r tually no relationship was found between job perfor mance

and job satisf action. The most appropriate answer to the dilemma of

the relation between production and job s a t is f a c t i o n app ears to have

been offered by Brayfield and Crocket t (1955), who described it as one of

conc omi tant variation rather than cause and effect. They assume that

74
~

individuals are motivated to achieve certain goals whose achi eve ment re-

suits in satisfaction. Productivity is seldom a goa l in i tself but is more

commonly a means to goal a ttainment. High satisfaction and high p roduc-

tion can be expec ted to occur together when productivity is perceived as

a means to certain important goals. Under other conditions the relation-

ship may be negative or there may be none at all. The possibili ty exists

that in the surveyed environment , productivity may not be perceived as

a means to certain important goals desired by the scientists and engineers.

Hypothesis 2. A positi ve rela ti onship exists between producti-

vity and an internal.

Results: This hypothesis applies to the tota l Rotter score since

-~

the measure of inte rna l /exte rna l by the I-E scale may be invalid. The
hypothe sis was not supported at the . 01 leve l for all productivi ty va riable s

(Table XI). In fact for one producti vi ty variable (PVARS), a lower Rotter

score was related to the nonproductive person nel (Table XIV). Die to

the uncertainty of the significance of the tota l Rotter score , no justi fi-

cation is provided for this result.

- Hypothesis a A positive relationship exists between job satis-

faction and an internal.

Results: A si gni fi cant relationship was found at the .01 level

with the total Rotter score (Table XXIV). This negative relationship

indicates that the hig her the job satisfaction, the more internal the indi-

vidual would be as measured by the I- E scale .

75

- _
 Prye r and Di stefano (1971) and Lichtma n (1970) also

found some signi ficant correla tions between the total Rotter score and

job satisfaction at some employee levels. Their findings and the present

results indica te tha t the total Rotte r score may be significantl y related

to job satisfac tion, but the concept which is measured by this score

re mains unde ter mined.

I-Ivpothesis 4. A positive relation ship exists be tween produ cti-

vi ty and job satisfaction and positive leader reward behavior. A nega ti ve

re lationshi p exists between both productivity and job satisfaction and

puni ti ve leader reward behavior.

Results: The first part of this hypothesis was not supported by

the results at the .0 1 leve l for any produc ti vi ty va riable s (Ta b le XI).

Addi tionally , no significant differe nce was determined between produc ti ve

and nonproduc tive pers onnel based on positive lea der reward behavior

(Table XIV) . However , for job sa ti sfaction , the first part of thi s hypo-

thesis was found to be significant at the . 01 level (Table XXIV).

Sims and S7i lagyi (1975) found a positi ve relation-

sMp between positive leader reward behavior and pe rformanc e evalua-

tion, however, their measure of performance indicated not only quanti ty

of work (productivity) , but also eight other factors. It can only be

conjectured , but this re sult suggests tha t the scientists /en gineers may

not be rewarded on the basis of productivity but on the basis of some

other measure of performance.

L .
-.
_ _
. - _ . __. I , *
~~~~~*
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~
 _ __ __ __ _ _
_ _

Si ms and S7ilagyi also found a positi ve relationship

between job satisfaction and pos i ti ve leader reward behavior as did this

study .

The secon d part of thi s hypothe sis was not supported

at the .01 level for any producti vi ty variables (Table XI) . In one case

(PVAR4). a nega tive re la ti onship exi sted at the . 05 level (Table XVI) .

No significant difference was de ter mi ned between productive and non-

produc ti ve personne l based on punitive leader reward behavior (Table

XIV) . Addi tionally, the second part of this hypothesis was not sup-

porte d at the .01 level by the job satisfaction results (Table XX IV) .

Sims and S7i lagyi (l~~ 5) reported several significant

negative re lationships between punitive reward be ha vior and per fo r mance ,

however, they reported no significant re la tionship for the pro fessional

and technical groups studied. Sinc e scientists /engineers would be con-

sidered professional/technical person ne l , the current results are con-

sistent with their findings.

Sims and S7i Ia gyi reported no significant relationships

qf job satisfaction with puni ti ve reward behavior.

t~ypothesis 5. A positive relationship exists between both pro-

ductivi ty and ~ob satisfaction and the supervisory style of “Conside ra tion ”.

Results: Thi s hypothesis was not supported at the .01 leve l for

all prod uctivity variables (Table XI ) . In one case (PVAR6 ) , a negative

relationship exists at the .05 level. Additionally , no sign ificant diffe r-

ence was determined between productive and nonproductive based on

“Considera tion ” (Table XIV).

77
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
~

Pelz (1952) found that prov

iding supportive leadership

(considera tion ) for the men is useless if the supervisor always fails to

persuade higher management. The result being less rather than more

productive employees. Whe the r this attempt at “consideration” is pre-

sent in the surve yed laboratory is unknown, but a message for lack of

association between “consideration ” and productivi ty may be present in

P~ 1z ’ s fi nding.

A significant positive relationship was foun d between

job satisfaction and “Considera tion ” at the .01 level (Table XX 1V).

This finding is consistent with the study of job satisfaction and conside-

ration by Baumgartel (1956; 1957) who found positive relation ships between

supportive leadership (consideration) and job satisfaction in researc h

laboratories.

Hypothesis 6 . A ne gative relationship exists between both pro-

ductivity and job satisfaction and the supervisory style of “Initiating

Structure ”.

Results: This hypothesis was not supported at the . 01 leve l for

all productivity variables (Table XI ) . In two ca ses (PVAR l and PVAR 2)

~
a negative relationship existed at the . 05 level. Additi ona lly. no signi-

ficant diffe rence was determine d between productive and nonp roductive

personnel based on “Initiating Structure” (Table XIV) .

Korma n (1966) revealed several studies showing no

relationship between i n i t i a t i n g stru ctu re and performance which sup-

ports most of these results.

78
 -- - . .-
- - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~ ~~

For the j ob satisfaction results , this hypothesis was

not supported at the . 01 level (Table XXIV ) . This resuit is also sup-
ported by the Korman fi ndings which revealed severa l studies showing

no re lationshi p between initiative structure and job sati sfaction.

4. 7. 2 Comparison for Group Leaders

The following hypotheses with the app licable results for the group

leaders are provi ded. In some ca ses, the justification or rationale for

the results are the sa me as for the nonsupe rvi sor y scientists /engineers.

In those cas es , the rea der is referred to the previou s section.

Hypothesis 1. A positi ve relationship exists between productivity

and job satisfa ction.

Results: Thi s hypothesis was not supported by the results at the

.01 leve l for any productivity va riables (Table X I ). Addi tionally, no sig-
nificant differenc e was de termined between productive and nonproductive

personnel based on job satisfaction (Table XIII) .

These results are consistent with the findings of Vroom

(1964) and Bra yfleld and Crockett (1955) discussed in the previous section.

Hypothesis 2. A positi ve relationship exists between producti-

vi ty and an internal.

Results: This hypothesis applies to the tota l Rotte r score since

the measure of inte rna l ‘exte rna l by the I-E scale may be invalid. The
hypothesis was not supported at the . 01 level for all productivi ty variables

(Table XI) . Additionally, no si gni ficant diffe rence was determined between

79
 product
ive and nonproductivepersonnel based on the total Rotter score

(Table XIII). Due to the uncertainty of the significance of the total Rotte r

score , no justi fication is provided for this result.

hypothesis 3. A positi ve relationship exists between job sati s-

faction and an internal.

Results: Thi s hypothesis was not supported at the . 01 level

(Table XI) . This result is in contrast to that determined for the non-

supervisory scientists /engineers whe re a significant relationship was

found between job satisfaction and the tota l Rotte r score. In their study

of job satisfaction and internal/ external Pryer and Distefano (1971) sug-

gested that different employee levels would revea l dissimilar r elationships

be tween internal ’exte r nal score s and job sa tisfac tion. The results of
their stud y and the current re sults are consistent wi th this suggestion.

~ ypothesis 4. A positive relationship exists between both pro -

ductivity and job satisfaction and posi tive leader reward beha vior. A
ne gative re lationshi p exists between both product ivity and job satisfaction

and puni tive leader reward beha vior.

-
Results: The first part of this hypothesis was not supported by

the results at the .01 leve l for any producti vity va riables (Ta b le X I).

Addi ti onally, except for PVAR 1, no signiticant di fference was determined

between productive and nonproductive personnel based on positive leader

reward behavior. For PV AR 1, the productive personnel perceived their

supe rvisors as exhibiting less positive leader reward behavior than did

80

- -. ~~~-.- .~~~ - -~~~~~~~~~~~~~~~~~ -

 ~~ -- -~~~~ ~~~~~~~~~ -~~~ . -— - -~~ ~~~~—~~~~~~~~~~~~~~~~~ - -• - • . -
— .

the nonproductive personnel at the .01 level (Table XIII ). However, for

job satisfaction the first part of this hypothesi s was found to be significant

at the .01 leve l (Table XXIV) .

These results are c onsistent with the findings by Si ms

and Szilagyi (1975) discus sed in the previou s section.

The second p art of this hypothesi s was not suppor ted by

the results at the .01 level for all productivi ty variables (Table XI) . In

two cases (PVAR3 and PVAR5), a positive relationship existed at the . 05

level (Table XI) . Additionally, except for PVAR3, no significant diffe r-

ence was determined between productive and nonp roduc ti ve personnel based

on punitive leade r rewa rd behavior. For PVAR3, the productive personne l

pe rceived their supe rvisors as exhibi ting more punitive leade r reward be-

havior than did the nonproductive personnel at the . 026 level (Table XIII) .

Furthermore , for job satisfaction the second part of this hypothesis was

not supported by the results at the .01 level (Table XXIV).

The lack of a negative association for professiona l and

technica l personnel is consistent with the findings of Sims and S~’i la gyi

(1975).
-
Hypothesis 5. A positive relationship exists between both pro-

ductivity, and job satisfaction and the supervisory style of “Consideration”.

Results: Thi s hypothesis was not supported at the . 01 leve l for

any productivity variables (Table XI) . Add i ti onally, no significant dif-

fe rence was determined between productive and nonp roductive personnel

based on “Consideration ” (Table XIII). However, a significant positive

re lation shi p was found between job satisfaction and “Consideration” at

th e . 01 level (Table XXIV) .

81
-----
~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - —-fl -~ ~~~
_ ____
-
 These results are consistent with the findings by
F~ 1z (1952) and Baumgarte l (1956; 1957) discussed in the previou s section .

hypothesis 6 . A negative re lationship exists between both pro-

ductivity and j ob sati sfaction and the supervisory style of “Initiating

Struc ture ” .

Results: This hypothesis was not supported at the .01 level for
all productivi ty variables (Table XI). However, one si gnificant differerx e

was determined betwee n p roductive and nonproductive personnel on PVAR2.

The productive personnel perceived their supervisors as exhibiting less

“Ini tiating Structure ” than did the nonproductive personnel at the .031

level (Table XIII) . No additional significant di ffe rences were determined

be tween producti ve and nonproducti ve personne l based on “Initiatin g

Struc ture ” (Table XIII). Furthermore, this hypothesis was not supporte d
by the results at the . 01 leve l for job satisfaction (Table XXIV) .

These results are consistent with the findings of Korma n

(1966) discussed in the previous section.

82
 CHAFFER 5

DISCUSSION AND CONCLUSION S

5. 1 Intr oduc tion

In Chapter 1, the following obj ective s were sta ted as guidance for

this researc h effort :

I) Using the developed survey, collec t data on a sample of

scientists /engineers from an AF R&D laboratory.

2) Test the relation ships among the following factors: produc-

tivity, job satisfa cti on , gene rali zed expectancies for internal versus

externa l control of reinfo rce ment, leade r reward behavior , supervisory

style , and certain stati stical control data such as educati on , age , experi-

ence, etc .

3) By using the collected da ta , a tte mp t to recommend certain

labora tory policies which would improve / maintain the current manage-

ment practices.

The firs t two objective s have been accomplished; the third objec ti ve

remains and it is the purpose of this chapter to accomp lish this obj ective

by stating the ma na gement implications of the tested hypotheses. Prior,

however , to discussing this subjec t with the shortcomings of this effort

and recommendations for further researc h , the validity of the measure-

ment scale s used in this study are discussed.

83
 - —- - . -- - -~~~~-—- ~~ - —- --
— ~~~ ~~~ ~~~~~~~~~~~~~~~~~~~~~

5. 2 Validity of Measurement Scale s

Four well-known measurement scales were used in this study :

I) the Supervisory Behavior Description Questionnaire:

2) the Leader Reward Behavior Instrument:

3) the Hoppoc k Job Satisfaction measure : and

4) the Rotter Internal-External Scale.

With the exception of the Rotter 1-E scale , these measures appeare d to

be valid and reliable.

As indicated in Chapter 4, factor anal ysis of the Rotte r I-E

scale resulted in seven factors. One fac tor resulted in the clear-cut
construct of politica l control wh ile the remaining factors were a com-

bi nation of persona l and ~zenera l control items .

It can only be conj ectured, but the c lear-cut Ioadin ~ of the amole
~
on onl y the political control factor may be an i ndication of the political

environ ment in whic h the personnel ope ra te or a high gene ra l concern

for politica l matters b y the respondents.

However, the existence of seven fa c tors sheds considerable

doub t on Rotte r ’s contention that the I-E scale is unidi mensional and

measures a general perception of control of reinforcement. In fact,

the I-E sca le is definitel y multi-di mensiona l for this sa mp le and , through

comparisons with studies by Gurin (1969) and Cherlin (1974). may be

sensitive to population differences .

84

L .. - — --~~-
~ -.- -- - . -- -— - -- ~~ —- -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ —~~~

In conclusion , future researchers should be awa re of the ques-

tionable va lidity of the Rotte r I-E sca le. r)u e to the obviou s multi -

di rnensionality of the scale , the question remains as to the nature of the

concept being measured.

5.3 Implications for Manage ment of the Tested l-Tvpothe se~

The na ture of the hypotheses reflected the desire to determine

supervisory and individua l variables which might be associated wi t -h pro-

ducti vi ty and job sa tisfaction. As with past studies, th e variab le s which

are associated wi th producti vi ty were not consistent . although the vari-

ab les associated with job satisfaction revealed some consistency .

No single variable was shown to be si gnificantly associated with

all six of the produc tivity variables. The variables iuenti fied in (‘ a t-ej-
~
4 highlighted this inconsistent pa ttern of association . The corre lation s

provided in Tables X IV and XV showe d a significant trend of higher edu-

cation , more yea rs of experience, and higher grade as being posi ti vely

associated with producti vi ty. These associated variables w’ uld gene-

rally be expec ted.

Table XVI , however , revealed some expected results and one

surp rising result. Initiating structure was found to be negatively asso-

ciated wi th the outputs of journa l articles and unpublished manuscripts.

This association may be du ~ to the time con straints associated with a

struc tured work situa tion. Additionall y, punitive leader reward behavHr
was negativel y associated wi th the outpu t of technic al memorandum.

85

_ _ _
AO—AO ’45 981 AIR FORCE IN5T OF TECH WR IGHT—PATTERSON AFU OHIO SCH—ETC FIG 5/9
PRODUCTIVITY AND JOB SATISFACTION IN RESEARCH AND DEVELOPMENT? —— ETC(IJ )
SEP Ti I. .J CORBIN
UNCLASSIFIED AFIT/GSM/SMfl75—2 NL

.
0 4 59 8 F _________________________________
_____________________________________________________________________ _
___________________
_______ _______________________________________
__________________

END
Dart
18.180
-77
Dec
 10 T28

~ llfl
1W 11
~ ~~
f ~~’4 ~f’i ”~’
“ ‘ 25 i.
flfflF

NAT IONAL BUREAU OF STANDAR DS

MICROCOPY RESOLUTION TE ST CHART
Since technical memorandum are basically an ind ividua l product , p oor

performance in thi s area may be dealt with b y more direct leader cor-

rective behavior.

Surprisingly , conside ra ti on was negati vely associated with the

output dealing wi th con tract monitoring. Thi s associati on may indica te

the use of the supervisor ’ s conside ra ti on as a cru tch for low produc ti vi ty

in this area.

The group leaders revealed two significant correlations betwee n

productivity and the predi ctor variables. Puni ti ve leader reward be-

ha vior was positively associated with technical reports and new or

improved processes, p roducts, techniques. and patents or pa tent appli-

cation. These associations indicate tha t group leaders perceive that thei r

supervisors will use punitive rewa rd behavior if the output in these cate-

gories is low. Whe ther this perception is due to laboratory policy or

not , the answer is unknown because the questionnaire did not record the

importance of these output s to the personne l or mana ge ment.

The lack of consistent association of produc ti vi ty variables to

predictor variables may be a function of the amount of rewards tha t are

conti ngent upon productivity in the Civil Service-Military system and/or

the a mount of control which a supervi sor has over rewards in this sys-

te rn. Additionally, both of the supervi sory measure s used in thi s

study were origi nally designed for use in civilian i ndustry where the

86

1
 ~

.
supervi sor tends to have more control over rewa rds such as pay and

pr omotion. The combination of these two factors may have li mited

the derived associati on s between variables.

in contrast to the productivity variables, two predictor variables

were positively rela ted to job satisfaction for both the nonsupervisorv

scientists /engineers and the group leade rs. Consideration and positive

leader reward beha vior were associated with higher job satisfactory

for both groups. These fi ndings are consistent with previou s stu dies
(House, 1971; Sims and S ’ilagyi , 1975) and represent the i ndividuals’
~
prefe rences for a supervisor who establishes a cli mate of good rappor t

and two-way communication and who also will reward g ood performance

with leader-admi nistered rewards.

In addition to these two predictor va riables, education was

associated with job satisfaction for both groups of personnel. Education

was positively associated with job satisfaction for nonsupervi sorv scien-

tists /engineers, possibly indicating an appropriate matc h of educational

background with the job. However, in the case of the group leaders,

education was negati vely associated job satisfa cti on. Thi s re sult may
indicate that those group lea ders with hi gher educ a ti ona l levels percei ve

their job s as not using thi s background of knowledge.

Finally, for the noisupervisory scientists /engineers, a negative

association exists between job satisfaction and the total Rotte r score.

Due to the questionable na ture of the concept represented by thi s score ,

no ju stificati on for thi s association can be p rovided.

87

.- . .
~~
1~~ ~~~~~~~~~~
•- -
~~-~~~~~ ~~ — —-
.
~ _ _

Thus, the scientists/engineers indicate a preference for super-

vi sors who are conside ra te and who will reward good performance.

Unfortu na tely, no comparable predictor variables are associated with

productivity.

5. 4 Implicati ons of Team Development

The team development progra m for the laborato ry was in its

infa ncy stage when the survey was conduc ted. However, the results

indicate a positive relationship between both the awa reness of and parti-

cipation in the progra m and job sati sfaction. The hi ghe r job satisfaction

could indicate the effectiveness of the progra m or indicate employee

satisfaction at the attention received from labora tory mana gement.

Unfortunately, the reason s for the highe r job satisfaction were unde ter-

minable from the two tea m development questions.

Since the aware ness question deals primarily with the communi-

cation of the program’s existence through laboratory channels, the low

pe rcenta ge of aware respondents may indicate a need to improve la b-

oratory communication processes.

Lastly, no diffe rences existed between p articipants and nonparti-

cipants based on productivity. It would be inva lid to conclude that the

team development progra m has had no effect on prod uctivi ty due to

I) the relatively short existence of the progra m and 2) the time peri od
associated with the productivity questions.

88

~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~ _ _ _ _ _ _ _ _ _
 5. 5 Shortcorning~ of Current Research

As mentioned by Stahl (1975) , resea rc h performed in an envi ron-

ment where the amount of rewa rds and the control of these rewa rds ba sed

upon performance is limited may have been a handicap in determining

variable relationships. Thus, the limi ted association of the supervisory

characteristics and productivity variables may ha ve been i nfluenced by

the Civil Service-Military system.

-

~.

The small number of respondents resulted in the inabili ty to fa cto r

analyze the SBDQ. Thus , the di mensionality of this measure in relation

to the sample was not determined.

The use of the Rotter I-E scale may have influenced the response

ra te. In the questionnaire section provided for respondent comments,

nume rous completed and uncompleted questionnai res were highly critical

of the intent and validity of the scale. Addi tionally, the Rotte r I-E scale

was the most frequently unanswered portion of the questionnaire with

approxi mately 17 per cent of the returned questionnaires being blank in

this section.

5. 6 Recommendations for Future Resea rch

The following suggestions for future resea rch are provided based

t~ on the shortcomings of this research effort and the researcher’s per-

ception of needed efforts.

Further study should concentrate on the va lidity and reliability

of the Rotter I-E scale. This scale is often used as a measure of the

89

L -
~~~~~ ~~~~~~~~~~
- - - - -- —
~~~- . .-
. - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~ - —

internal/externa l concept and is employed as a valid instru ment. The

current study shows the scale to be invali d however, due to the emphasis

on and the implications of the internal/e xterna l concept, additional re-

search should continue the investigation of the Rotter I-E scale.

Due to the possible influence of the Civil Service-Militarysystem

on the reward concepts measured in thi s study , this study should be repli-

cated in industrial and uni versity research and develop ment labora tories.

Such research would provide additional i nformation on the relation of

rewards to productivity and job satisfaction.

The influenc e of the tea m development progra m should be incor-

pora ted into a longi tudinal study . This study should focus on the effects

of team development programs on job satisfaction and producti vi ty ove r

an extended ti me fra me.

The comments provided by the respondents and the results of the

awareness team development question both indicate a need for the study

of intra lahora tory communications as it relates to prod uctivity and job

sa tisfacti on.

Finally, a study of additional organi zational variables such as job

challenge and autonomy shou ld be incorporated into futu re studies of gov-

ernment researc h and development laboratories as they relate to pro-

ductivity and job sati sfaction. As indicated by respondents comments,

there was widespread concern about job con ten t and di fficulty . These

job characteristics may prove to be related more to producti vi ty than

were the supe rvi sory cha racteristics.

90

_ _ _ _ _.
_
~~~~~~~~
 _

BIBLIOGRAPHY

Baumgartel , H. “Leadership, Motivations , and Attitudes in Resea rch

Laboratories”. J~~rna l of Social Issues. 12: 24-31 (1956).

Baumgartel, H. “Leadershi p Style as a Variable in Resea rch Admini-

stration”. Admi ni strative Science Qua rterly. —
2: 344-360
(Dec 1957).

Bradsha w , H. H. “Need Satisfaction, Mana gement Style. and Job Level

in a Professiona l Hierarchy ”. ~~ peri mental Publication System.
8: Ms. #289-1 (1970).

Brayfield , A. H. and W. H. Crockett. “Employee Atti tudes and Employee

F~ rfor niance”. Psychological Bulletin. 52: 396-424 (1955).

Cherlin , A. and L. B. Bourque. “Dimensionality and Reliability of the

Rotter I-E Scale ”. Sociometry. 37(4): 565-582 (1974).

Evans, M. G. “Extensions To A Path-Goal Theory of Motivation ”. Un-

published pa per. The Ohio Sta te Uni versity, 1973.

Fournet, G. P., Distefano, Jr. , M. K . , and M. W. Pryer. “Job Satis-

facti on : Issues and Problems ”. P~ rsonnel Psychology. 19(2):
165-183 (1966).

Franklin, R. D. “Youth’s Expectancies About Internal Versus External

Control of Reinforcemen t Related to N Variables ”. Dissertati on
Abstracts, 1963, 24 , 1684.
—

Freund , J. E. Mathema tical Statistics (2nd Ed.). Englewood Cliffs:

Prentice-Hall, Inc. , 1971.

Gurin , P. , Gur in , G. , Lao, R. C., and M. Beattie. “Internal-External

- Control in the Motivational Dynamics of Negro Youth”. Journal
of Social Issues. 25: 29-53 (1969).

Harma n , H. H. Modern Factor Analysis. Chicago: The Universi ty of

Chicago Press, 1976.

Hill , J. W. , and J. G. Hunt. “Managerial Level, Leadership, and

Employee Need Satisfacti on ”. In: Current Developments in the
Study of Leadership! Carbondale: ~~~ uthern illinoi s Universi ty
~~~ ss, 1973.

91

~~~~~~~~~ ~~~~~~~~ ~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~

~ .- . . .- - -
-. .- —— .- - .
-
- ~~- .- - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

House, R. J. Filley, A. C., and S. Kerr. “Relation of Leader Consider-

ation and Ini tiating Structure to R and D Subordi nates ’ Satisfaction”.
Administrative Science Quarterly. 16: 19-30 (1971).

Hughes Aircra ft Company . R&D Productivity : An Investigation of Ways

to Evaluate and Impro~ë Productivityin Research and Development.
~T’he results of a study performea by Hughes Aircraft Company,
September 1974.

Joe , V. C. “Review of the Inte rnal-External Con trol Construct as a

F~ rsonality Variable”. Psychological Report s. 28: 619-640 (1971).

Johnson , R. D. “An Investigation of the Interaction Effects of Ability and

Motivational Variables on Task Performance ”. Unpub li shed disser-
tati on. Indiana Uni versity , 1973.

Keller, R. T ., and A. D. Szilaygi . “A Comparative In vestiga ti on of the

Supervisory Behavior Description Questionnaire (S BDQ) and the
Revised Leader Behavior Description Questionnaire (LBDQ- Form
XII)” . Academy of Management Jou rnal. 19(4): 642-649 (Dec 1976).

Kerr , S., Schrieshei m , C. A. , Murphy , C. 3. and R. M. Stogdill.

“Towa rd a Contingency Theory of Leadership Based Upon the Con-
sidera tion and Ini tiating Structure Literature ” . Q~gani zationa 1
Behavior and Huma n Performa nce. 12: 62-8 2 (1974).

K orman, A. K. “Consideration ”, “Ini tiating Struc ture” , and “Organi-

zationa l Criteria-A Review ”. Personnel Psychology. 19(4):
349-361 (Winter 1966).

Lawler, E. B. a rid L. W. Porter. “Antecedent Attitudes of Effective

Managerial Performance ”. Orga ni z~ tionaI Beha vior and Hu man
Performa nce. 2: 122-14 2 (1967).

Lawler, E. B. and D. T. Hall. “Relationship of Job Characteristics to

-
Job Involve ment, Satisfacti on, and Intrinsic Motivation”. lournal
of Applied Psychology~ 54(4): 305-312 (1970).

Lefcourt , H. M. “Interna l Versus Externa l Control of Reinforcement:

A Review ”. ~~ychologica 1Bulletin. ~~~4) : 206-220 (1966).

Lichtrna n, C. M. “Some Intrapersonal Re.;ponse Correla tes of Organi-

zational Rank”. J~~rnaI of Applied Psychology. 54: 77-80 (1970).

92

.—
~- “~~~~ ~~~~~~~~~ .
 _ _ _ _ _ _ - - . - .-
_ _ _ _ _ _ _

McCarrey, M. W. and S. A. Edwa rds, “Orga nizationa l Cli mate Conditions

for Effective Research Scientist Role Performance ”. ganizationa i
Behavi or and Hu man Performa nce. 9: 439-459 (1973).~~

McNlchols, C. W ., Stahl , M. J ., and T. R. Manley. “The Validity and

Reliability of Hoppock’ s Job Satisfaction Measure”. Unpublished
paper. Wright-Patterson Air Force Base: Air Force Institute of
Technology, 1976.

Morrison D. G. “On the Inte rp retation of Discrirninant Anal ysis ” . Journal

of Marketing Research. 6: 156- 163 (May 1969).
Nie, N. H., Hull , C. H., Jenkins, J. G., Steinbrenner, K . , and D. H.
Bent. Statistical Package for the Social Sciences (2nd Ed. ). New
York: McGraw-Hi1T~ 1975.

FèIz, D. C. “Influence: A Key to Effec tive Leadership in the First- Line H

Supervisor”. Personnel. 29: 209-217 (1952).
Pelz, D. C. “Some Social Fac tors Rela ted to Performance in a Resea rch
Organi za tion ”. Administrative Science Quarterly . 1: 310-325 (1956).

Pryer, M. W. and M. K. Distefano, J r. “Perceptions of Leadership Be-

havior , Job Sati sfa ction, and Internal-External Control Across
Three Nursi ng Levels”. Nursing Research. 20(6): 534-537 (1971).

Rotter, J. B. “Generalized Expectancies for Interna l Versus External

Control of Reinforcement ”. Psycho1o ical Monographs. 80(1):
260-294 (1966). ~

Si ms , H. P., Jr. and A. D. Szila gyi. “ Leader Reward Behavior and Sub-
ordinate Satisfa cti on and Performa nce”. Or aniza tiona1 Behavior
and Human Performa nce. 14: 426-438 (1975). ~

Stahl , M. J. “Innovation and Productivity in Research and Develop ment:

-
Some Associated Indi vidua l and Organi zationa l Variables”.
Unpublished dissertation. Rensselaer Polytechnic Ins titu te :
Troy, New York (Decembe r , 1975).

Stevens , A. E. “Rewa rds for Air Force Research and Developmen t

Scientists /Engi neers”. Unpubli shed thesi s. Wright- Patterson
Air Force Base: Air Force Institute of Technology , 1976.

Stogdill , R. M. and A. J~ . Coons. Leader Behavior: Its Description and

Measuremen t. Columbus: iii eau of Business Researc h, Ohio
~ ~
State Uni versity, 1957.

93
 —— -—
— . - -
-.-.—
~~ ~~ ~~~~~~~~~~~~~~~

Vincent, H. F. and A. Mirakhor. “Relationship Between Productivity,

Satisfaction, Abili ty, Age , and Salary in a Milita ry R&D Organi-
zation”. IEEE Transactions on Engineering Mana gemen t, EM-19:
45-53 (Ma~7i~72).

Vroom, V. H. “Ego-Involvement, Job Satisfaction , and Job Performance” .

Personnel Ps ychology. 15: 159-177 (1962).

Vroom, V. H. Work and Motivation. New York: Wi ley & Sons, 1964.

SUPPLE ME NTAL BIBLIOGRAPHY

Draper, N. R. and H. Smith. Applied Regression Analysis. New York:
John Wi ley and Sons, Inc. , 1966.

Siegal , S. Nonparame tric Statistics. New York : McGraw-Hill, 1956.

Tatsuoka , M. M. Multivariate Analysis: Techniques for Educational and

Psychologicil Research. New York: John Wiley and Sons, Inc. , 1971,

94

L - --
-- ,
~~~~ ~~ ~~~ ~~~ ~~~~~~~~~~~~~~
. — .,,. .-
.-
~~~~~~~~~~~~~~~~~ ~~ --
_ _ _ _ _ _ _
~~~~ - - . ---. -~~~ -~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _ _ _

AP~~ NDIX A I

QUE STIONNAIRE

95 I

-. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ — - — — -~~~~~~~ .
~~~~ - .- -., : ~~~~~~~~~~~~~~~~~~~~~~~~~~~
—
I
 --. - -
~~

.
O!~~ART M E~~T CF T~-~E A~~ FCRCE
C
A’R F O RC E ~~~~~~~~ ~~V ’ ~ . . M t C S LA ~~ OR~~~~3~~ ’~ —‘
,v ~~c z :~~ ..3E 4 54 i 3
~~~It I i -P A i~~~~~~~~~~~~ ~ ~ ~~~~~~~~

•i i 1 . 4

~~ CC ~~~~~~~ ~
A Study oz Pro ductivity and Job Sat~ sIaction in
the AF R&D Laboratories (Charge Nr: 99949801 )
-~~ AFFDL 3cienti st/ ~ng ir.eers

1. Cap t Larry Corbin , an A~ IT student , is cond~ cting a

survey of the R&D Laboratories by mean s of the attached
questionnaire . The data from this survey should help
to identify those individual and organizational factors
which impact upon productivity and job satisfaction .
Capt Corbin needs your help in collecting this data
through your timely and thorough completion of this
questionnaire .
2. The envelope provided should retu~~ your answers
directly to Capt Corbin through base distri’~~:iQr. ar.d
anonymity shculd be iai a~ned t o~~ ho’~t ~~~ i~~~~t~.-
~-a:ion . I~ there are any :~
Oa t C rb n a ;
uescions, please c~ r.:act
ary o~ ~he survey re~~~ ts
~
.iill ~ ~~~~~
~ provided
~ ~~ .
~~

be to all erested res ~ onden~~ .

3. I believe we 1can al]. profit from a study of tnis typ e
and your assistance in completing the survey is requested.

ALE~ RT E. PREYSS, lonel, ~SAF 1 Atch

Commander Questionnaire
w/return enevelope
—

96

-. ~~~~~~~~~~~~~~~~~~~~~~~
.
 —
_.
~~~~~~~~ —— ~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Dear Survey Participant,

I am a sking for your volun t ary cooperation in a study con-
cerned with productivity and job satisfaction in Air Force R&D
Laboratories.
Your res p onses w ill be kept in the str ictes t confi d ence an d
your anonymity is guaranteed . All questionnaire data will be
reported in aggregate form only. Your name was derived from a
roster of all AFFDL scientists/engineers and there will be no
attempt to relate your name to the completed questionnaire .
When you have completed your q ues tionn aire, place i t in the
attached self-addressed envelope , seal the envelope , and pl&ce
the envelope in Base distribution. I will provide a summary of
the results of my survey to any interested respondents . If such
a summary is desire d , p le ase sen d a reques t to me un der se para te
cover to Capt Larry Corbin , AlIT/ENS, Wright-Patterson APE , OH
45433 through Base distribution (this method will assure anonymity) .
The number in the bottom left corner of this survey is a
USAF contro l number only and is required on all surveys within
the AF. The sam e number is on all questionnaires and in no can
it be used to identify the respondent.
Thi s s tu dy is b ein g con duc te d in p ar tial fu l f i l l m en t o f the
r equ ir emen ts for my mas ter ’ s degre e in Sys tems Man agemen t w ith
AFIT. Your cooperation in completing and returning this question-
naire by 1 May 1977 is vital to the completion of this study . The
data and analysis for this study is totally dependent on the
quantity, quality, and completeness of your responses. The quan-
tity of questionnaires is as important as the quality of responses
in ord er to p rovi de a sufficien t samp le size of scien tis ts/en gi-
neers within AFFDL to draw valid conclusions about productivity
and job satisfaction .

- I apolo g ize for the amoun t of time and e f f o r t r equire d to

complete this questionnaire , but I feel , as Colone l Preyss does ,
tha t this s tudy can b e of benef it to all scien tis ts/en gine ers
and your time and labor is greatly appreciated. Thank you again
for your valuable assistance .

LARRY .XJC~ 4BIN , Capt , USAF

Graduate Student in Systems Management
Air Force Institute of Technology

USAF SCN 77-77

97

.
.
.TT ITIEJI.
~ .
_

PRIVACY STATEMENT

In accordance with paragraph 30, APR 12—35 , the following

information is provided as required by the Privacy Act of 1974:
a. Authority
(1) 5 U.S.C . 301, De par tmen tal Re gula tions:
and/or

(2) 10 U.S.C. 80—12 , Secretary of the Air Force,

Powers and Duties, Delegation ~~~~~ .

b. Principal purposes. The survey is being conducted to

collect information to be used in research aimed at illuminated
and providing inputs to the solution of problems of interest to
the Air Force and/or DOD .

c. Routine Uses. The survey data will be converted to

information for use in research of management related problems .
Results of the research based on the data provided will be
included in written master ’ s theses and may also be included in
published articles , re por ts , or texts . Distribution of the
resul ts of the rese arch, ba se d on the survey data , whe ther
wri tten form or orally presen te d , will be unlimited.
d . Participation in this survey is entirely voluntary .
e. No adverse action of any kind may be taken against any
indivi dual who elec ts no t to par ticipa te in any or all of this
survey .

J.

98

_ .- -
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~
~

MEASUREMENTS OF PRODUCT IVITY AND RELA TED I NDI VIDUAL

VARIAB LES IN THE AIR FORCE FLIGHT DYNAMI CS LABORATO R Y

PA.RT A

INSTRUCTIONS: P lease complete the following :

A. Age (in years using your last birthday ) _ _ _ _ _ _ _

B. Current Grade (GS level or Military grade) _ _ _ _ _ _ _

C. Years in Current Section _ _ _ _ _ _ _

D. Years under immediate supervisor (i.e. Group Leader

or Branch Chief (round to nearest year) _ _ _ _ _ _ _

E. Educational Level (Circle highest degree) B .S.

M.S. M.S. + Ph.D. Other (please speci fy ) _ _ _ _ _ _ _

F. Nature of Current Work (% of time per category) : ADPO _ _ _ _ _

Research ,
_ _ _ _
Development Engineering ,
_ _ _ _
Systems Program

Of f i c e Support ,
_ _ _ _
Supervisory ,

Other (please specify ) -

G. Percent of total working time contract monitoring _ _ _ _ _ _ _

H. Years of Scientific/Engineering experience (research ,

consulting , etc.) since f i r s t degree (nearest year ) _ _ _ _ _ _ _

I. How many of the following have you authored , presented ,

prepared , e t c . , over the past two years?
1. Published papers in professional/technical j ournals _ _ _ _ _ _

2. Unpublished manuscripts _ _ _ _ _ _ _

3. Technical Repor ts _ _ _ _ _ _ _ _

4. Technical Memorandums _ _ _ _ _ _ _ _

5. New or improved processes , products , and techniques,

and patents or patent applications _ _ _ _ _ _ _ _

6. Hardware/software specifications , test reports ,

test plans , statements of work , requests for
proposals _ _ _ _ _ _ _ _ _

7. Oral presentations to technical or professional

audiences _ _ _ _ _ _ _ _ _

J. Are you a group leader?

USAF SCN 77—77 99
_ _ :
. ..-- .--—— -- - .- — _ .
. . _ :
 ..-

PART B

INSTRUCTIONS : This is a questionnaire to find out the way in

which certain important events in our society affect different
people. Each item consists of a pair of alternatives lettered
a or b. Circle the letter which corresponds to the statement which
you more strongly believe to be the case as far as you ’re concerned.
Try to respond to each item independently when making your choice .
There are no righ t or wrong answers : this is a measure of personal $
belief.
Item

la. Children get into trouble because their parents punish them
too much .
b. The trouble ,iith most children nowadays is that their parents
are too easy with them.

2a. Many of the unhappy things in people ’s lives are partly due to
bad luck .
b. People ’ s m i s f o r t u n e s result from the mistakes they make .

3a. One of the major reasons why we have wars is because people
don ’t take enough interest in politics .
b. There w i l l a lways be wars , no matter how hard people try to
prevent them.

4a. In the long run people get the respect they deserve in this
world.
b. U n f o r t u n a t e l y, an individual’s worth often passes unrecognized
no matter how hard he tries.
5a. Th e idea that teachers are unfair to students is nonsense.
b. Most students don ’t reali ze the extent to which their grades
are influenced by accidental happenings .
6a. Without the right breaks one cannot be an effective leader .
b. Capable people who fail to become leaders have not taken
advantage of their opportunities .
7a. No matter how hard you try some people just don ’t like you.
b. People who can ’t get others to like them don ’t ‘..riderstand
how to get along with others.
8a. Heredity plays the major role in determining one ’s personality .
b. It is one ’s experiences irt life which determines what they’re
like.
9a. I have often found that what is going to happen will happen .
b. Trusting to fate has never turned out as well for me as making
a decision to take a definite course of action .
2

100
_ _ _ _ _ _ _ _ _ _ _ _ _
~~~~~~~~~~~~~~~~~~~~~~~~~~~~

lOa. in the case of the well prepared student there is rarely

if ever such a thing as an u n f a i r test.
b. Many times exam questions tend to be so unrelated to course
work that studying is really useless .

ha. Becoming a success is a ma tter of hard work , luck has li ttle

or nothing to do with it.
b. Getting a good job depends mainly on be ing in the righ t
place at the right time .
12a. The average citizen can ‘~ave an influence in government
decisions .
b. This world is run by the few people in power , and there is
not much the little guy can do about it.
13a. When I make plans , I am almost certain that I can make t
hem
work.
b. It is not always wise to plan too far ahead because many
things turn out to be a matter of good or bad fortune any how .
14a. There are certain peop le who are just no good.
b. There is some good in everybody .
15a. In my case ge tting wha t I wan t has little or nothing to do
with luck.
b. Many times we might just as well decide what to do by flip—
ping a coin.
16a. Who gets to be the boss of ten depen ds on who was lucky
enough to be in the right place first.
b. Getting people to do the right thing depends upon ability ,
luck has little or nothing to do with it.
17a. As far as world affairs are concerned, most of us are
victims of forces we can neither understand , nor control.
b. By taking an active part in the political and social affairs
the people control world events .
18a. Most people don ’t realize the extent to which their lives
-
are controlled by accidental happenings.
b. There really is rio such thing as “luck” .
19a. One should always be willing to admit mistakes.
b. It is usually best to cover up one ’s mistakes .
20a. It is hard to know whether or not a person really likes
you .
b. How many friends you have depends upon how nice a person
you are.
21a. In the long run the bad things that happen to us are balanced
by the good ones.
b. Most misfortunes are the result of lack of ability , ignorance ,
laziness , or all three .

lot
.4
22a. With enough effort we can wipe out political corruption .
b . It is difficult for people to have much control over the
things politicians do in office.
23a.Sometimes I can ’t understand how teachers arrive at the
grades they give.
b. There is a direct connection between how hard I study and
the grades I get.
24a. A good leader expects people to decide for themselves what
they should do.
b. A good header makes it clear to everybody what their jobs
are.
25a. Many times I feel that I have little influence over the
things that happen to me.
b. It is impossible for me to believe that change or luck
plays an important role in my life.
26a. People are lonely because they don ’t try to be friendly .
b. There ’s not much use in trying too hard to please people ,
if they like you , they hike you.
27a. There is too much emphasis on athletics in high school.
b. Team sports are an excellent way to build character .
28a . What happens to me is my own doing.
b. Sometimes I feel that I don ’t have enough control over
the direction my life is taking .
29a. Most of the time I can ’t understand why politicians behave
the way they do.
b. In the long run the people are responsible for bad govern-
ment on a national as well as on a local level.

102

., — ~~~~~~~~
 - , -- -- .
~~ ~~~~~~~~~~~~~~~~~

PART C

INSTRUCTIONS : The following questions deal with your perceptions

of your job. Select and circle the n umber of the answer which
you feel best applies to your job.
A. Which one of the following shows how much of the time you
feel satisfied with your job?
1. Never.
2. Seldom .
3. Occasionally .
4. About half of the time.
5. A good deal of the time .
6. Most of the time.
7. All of the time.
B. Choose the one of the following statements which best tells
how well you like your job .
1. I hate i t .
2. I dislike it.
3. I don !t like it. .~

4. I am indifferent to it.
5. I like it.
6. I am enthusiastic about it.
7. I love it.
C. Which one of the following best tells how you feel about
changing your job?
1. I would quit this job at once if I could .
2. I would take almost any other job in which I could earn
as much as I am earning now .
3. I would like to change both my job and my occupation .
4. I would like to exchange my present job for another one .
5. I am not eager to change my job , but I would do so if
I could get a better job.
6. I cannot think of any jobs for which I would exchange.
7. I would not exchange my job for any other.
D. Which one of the following shows how you think you compare
with other people?
1. No one dislikes his job more than I dislike mine .
2. I dislike my job much more than most people dislike theirs .
3. I dislike my job more than most people dislike theirs .
4. I like my job about as well as most people like theirs.
5. I like my job better than most people like theirs.
6. I hike my job much better than most people like theirs.
7. No one likes his job better than I like mine .

103
 - .-— —--
.. .—- - , —.
.-- -— . ..----- -— ,
-. .-.--- .
—.
—.-
~~~ — --
...
-.
-
-. -—-.
.
. -- -~~~~~~~~ .— ~~~~~ ~~ ~~

PART D

INSTRUCTIONS : Please complete the following by placing the

appropriate letter from the key on the line provided which best
describes your immediate supervisor ’ s behavior (i . e . , group
leader or branc~~chief).

A. Always B. Often C. Occasionally D. Seldom E. Never

_ _ _ _ _
1. He refuses to give in when people disagree with him.

_ _ _ _ _
2. He does personal favors for the subordinates under him .

_ _ _ _ _
3. He encourages overtime work.
4. He expresses appreciation when one of us does a good
job .
_ _ _ _ _

_ _ _ _ _
5. He is easy to understand.

_ _ _ _ _
6. He tries out his new ideas .

_ _ _ _ _
7.He demands more tha .~ w~ ran do.

_ _ _ _ _
8. He rules with an iron hand .

_ _ _ _ _
9. He helps his subordinates with their personal problems .

_ _ _ _ _
10. He criticizes poor work .

_ _ _ _ _
11. He criticizes his subordinates in front of others.

_ _ _ _ _
12. He talks about how much should be done.
13. He stands up for his subordinates even though it
makes him unpopular .
_ _ _ _ _

14. He insists that everything be done his way .

15. He encourages slow—working subordinates to greater
effort.
_ _ _ _ _

16. He sees that a subordinate is rewarded for a job

well done.
_ _ _ _ _

17. He waits for his subordinates to push new ideas

before he does.
_ _ _ _ _

_ _ _ _ _
18. He rejects suggestions for changes.

104

_ _ _ _ _ _ _
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~
~ - - -— --- .
~~~
.- - --

KEY
A. Always B . Often C. Occasionally D. Seldom E. Never

_ _ _ _
19. He assigns people under him to particular tasks.
20. He asks for sacrifices from his subordinates for the
good of the entire department .
_ _ _ _

21. He changes the duties of peop le under him without

first talking it over with them.
_ _ _ _

22. He insists that his subordinates follow standard ways

of doing things in every detail.
_ _ _ _

23. He treats people under him without considering their

feelings.
_ _ _ _

24. He sees to it that people under him are working up to

their limits .
_ _ _ _

25. He tries to keep subordinates under him in good stan—

ding with those in higher authority .
_ _ _ _

_ _ _ _
26. He offers new approaches to problems .

_ _ _ _
27. He resists changes in ways of doing things.

_ _ _ _
28. He “rides ” the subordinate who makes a mistake .
29. He insists that he be informed on decisions made by
subordinates under him.
_ _ _ _

_ _ _ _
30. He refuses to explain his actions .

_ _ _ _
31. He lets others do their work the way they think best.

_ _ _ _
32. He acts without consulting his subordinates first.

_ _ _ _
33. He stresses being ahead of competing work groups.
34. He stresses the importance of high morale among those
under him .
_ _ _ _

_ _ _ _
35. He “needles” subordinates under him for greater effort.

—
36. He backs up his subordinates in their actions .

_ _ _ _
37. He is slow to accept new ideas.
38. He decides in detail what shall be done and how it
shall be done.
_ _ _ _

_ _ _ _
39. He treats all his subordinates as his equal.
105
 KEY
A. Always B. Often C. Occasionally D. Seldom E. Never
____ 40. He emphasizes meeting of deadlines .
41. He criticizes a specific act rather than a particular
individual.
_ _ _ _

_ _ _ _
42. He is willing to make changes.
43. He makes those under him feel at ease when talking
with him .
_ _ _ _

44. He asks subordinates who have slow groups to get more

out of their groups.
_ _ _ _

_ _ _ _
45. He is friendly and can be easily approached.

_ _ _ _
46. He emphasizes the quantity of work.
47. He puts suggestions that are made by subordinates
under him into operation .
_ _ _ _

48. He gets the approval of his subordinates on important

matters before going ahead.
_ _ _ _

Please use the following key to rate your immediate supervisor ’s

use of rewards :
A. B .C D... ...E F G
Strongly Undecided Strongly Agree
Disagree
1. Your supervisor would personally pay you a compliment
if you did outstanding work.
_ _ _ _

2. You would receive a reprimand from your supervisor if

you were late in coming to work .
_ _ _ _

3. Your supervisor would lend a sympathetic ear if you

had a complaint.
_ _ _ _

4. Your supervisor would be very much aware of it if

there was a temporary change in the quality of your
_ _ _ _

work.
5. Your supervisor would recommend that you should be
dismissed if you were absent for several days without
_ _ _ _

notifying the organization or without a reasonable

excuse.

106
.- ..
I. ~ - -.-———.- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -
~ - .— -—..
.- — —---..- --
-~~~~~~ -~~ — ~~~~~~~~~

A . B . C .D •E •F •G
Strongly Undecided Strongly agree
Dis a gr ee

____ 6. Your supervisor would see that you will eventually go

as far as you would like to go in this organization ,
if work is consistently above average.

7. Your supervisor would get on you if your work was not

as good as the work of others in your department.
_ _ _ _

8. Your supervisor would recommend that you be promoted

if your work was better than others who were otherwise
_ _ _ _

equally qualified.
9. Your supervisor would help you get a transfer if you
asked for one .
_ _ _ _

10. Your supervisor would tell his/her boss if your work

was outstanding .
_ _ _ _

11. Your supervisor would give you a reprimand (written

or verbally ) if your work was consistently below accept-
_ _ _ _

able standards.
12. Your supervisor would recommend that you get no pay
increase if your work was below standard.
_ _ _ _

13. Your supervisor would show a great deal of interest if

you suggested a new and better way of doing things.
_ _ _ _

14. Your supervisor would give you special recognition if

your work performance was especially good.
_ _ _ _

15. Your supervisor would do all he/she could to help you

if you were having problems in your work .
_ _ _ _

16. Your supervisor ’s recommendation for a pay increase for

you would be consistent with his/her evaluation of your
_ _ _ _

performance.

—
17. Your supervisor would recommend that you not be promoted
to a higher level job if your performance was only
average .
18. Your supervisor , would encourage you to do better if
your performance was acceptable but well below what
_ _ _ _

you were capable of.

19. Your supervisor would recommend additional training or
schooling if it would help your job performance.
_ _ _ _

107

——--—
~

KEY
A......B • .C. .D
... . • E • F •G
Strongly Undecided Strongly agree
disagree
20. Your supervisor ’s evaluation of your performance would
be in agreement with your own evaluation of your per-
_ _ _ _

formance.
21. Your supervisor would increase your job responsibilities
if you were performing well in your job .
_ _ _ _

22. Your supervisor would always give you feedback on how

your work affects the total service of the organization.
_ _ _ _

PART E
INSTRUCTIONS : Please answer the following que stions ~~~ or no.
1. Are you aware of the organizational development or team
deve lopment program currently being conducted in the
laboratory?

2. Have you participated in this program?

Your cooperation in completing and returning this questionnaire

is greatly appreciated . The remaining space is provided for your
commen ts regarding the questionnaire (scope, length, intent, etc.).

108

-~~~~~~~~
 .-~~~~~~- - .

APPENDIX B

VARIABLE CODING SCHEME

109

~~~~~~ — -~~~.— ~~~~~~~ -- .—— . — —-~~~~~~~~~~ . — ..,- ~~—— . .— - -— ~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --

~~~-~~~~~~~~~ . ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
— -—
- - - —
~~~
- ----- -- -- -- -
~~~~~~~~~~~~~
-- -
~
-- ---
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-
~~~~~~

VARIABL ~ CODING SCHEME

Variab le
Abbreviation Value Code
AGE Actual Years Actua l Yea rs
GRADE GS 6 , 7 , 8 or 2nd Lt 1
GS9 , 10, or lst L.t 2
GS I1 3
GS l2 or Capt 4
G5 l3 or Maj or 5
GS 14 or Lt Colone l 6
GS 15 or Colone l 7
MILCIV Civi lian 0
Military 1
YRSSEC Actual Years Actual Years
YRSIS Actual Years Actual Years
EDCTN B.S. I
2
M.S. 3
NI .S. + 4
Ph. D. 5
NCWR Percentage of Total Work Time Actual Percentage
NCWD Percentage of Total Work Ti me Actual Percentage
CONM Percentage of Total Work Time Actual Percentage
YEXP Actual Years Actual Years
WAR J Actual Number Actual Number
PV AR2 Actual Number Actual Number
PVAR 3 Actual Number Actual Number
PVAR4 Actual Number Actual Number
PV AR 5 Actual Nu mber Ac tual Numbe r
PVAR6 Actual Number Actual Number
~VAR 7 Actual Nu mber Ac tual Nu mber
GPC Yes
No 0
RI to R29 Internal Answer 0
Externa l Answer 1
HI to H4 Scale Value I thru 7
Si to S48 Scale Value 0 thru 4
LI to L22 Scale Value 1 thru 7
AWARE Yes I
No 0
PART Yes I
No 0

110
 ..-- ~~~-.-~~~~~ -- .--~~~—— - .~~ —-
~~~
.--- .- .

,
‘
I’

APPENDIX C

CORRELAT ION MATRIX

FOR
R OYI’ER QUESTION S

111

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 ——
ott~ o
— —— ——
c c

——
O~ ~~ ’
~~~~~ .0 ~~~~ ‘.0 C
~~ ‘ C Lt’~ C’ ~‘4 ‘.0 Lr ~ C’. — N ir .0 .
C~
’J C’1
~~ ‘~~ ~ C’ ‘ C” ~~ ‘.0 CM c~~ ~~ ~~~
— .
C C C C 0 c’.~ CN. ~~
CM
.
— C” ~ CM CM — C’~ CM CM C’~ C~ CM C”~ C C
.
c~~ . •

~~ ~~‘ tn ‘.0 0’. CM N — CM Lt

~ ~~~
.
‘.0~~ ‘.
~~ C”4 Q

N C
CM ‘ ‘
~~ C— ‘.0 cr~ cY ~ c
~ CM CM
~~
CM ‘.0 0’.

—
—

——
cc N N ON
C’~ ‘0 N cc — C’ C
N_C __ . ~~ 0 C N
— CM LI
c ~
L1 tI 0’ 0’ C
~ C ~
c C CN ~~ U NC
CM
~~~~ ~~ ~~
—
cc
cc cV

— — ——
C N tI~

—
— C C~~ 0’ cc C’. CM N ‘.0 .‘0 ~~ C’ C’
I-
—
— cr’. — c’~ c”.~ c
0 0
2:
0
if’ ~~
- — — . . .
N
N N 00 Lf~ LO tt CM ‘.0 .‘0
~~‘ .‘0 0
cc cc .‘0 ~~~~ ‘.0 N 0 0’
N — 0 ’ u~ 0’ ~~ CM
~~

H
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I —

C ~~ Q

———
— CM u cr C c’ — — C
N 00 N ‘ ‘.
0C ~ ‘ .0 C~ 0’ N
‘0 ‘
-~~
—
—
. —~~~
— C CM —
~~
cY ~
~~
c
~
~~
— — ~~
‘- .
— C C

tt) — 0 0 ’ C ‘C cc cv c N ‘C c’ C 2 2
~~~ U 0 ’ . ’ C ’ C 0 ’ 0 ’ L (~)C
~ C M NN~
~~ ~~~~~~

-
~
0~~0’

C ~~~ — ‘
~~~ CM 00 00 00 N 00 0 0 0
CM C -.~ C N N N o CM ‘C C~ 00 N C
~
• •.
-~

mm

‘.0 S 0’.— — — — — — — —
mm
C— CM 0_ . N 0 0C . C M C ’ U) ,
CM
~~ if’ ‘.0 N 0’~~~~—~~~ CM cr ~ ~~~‘ Lt~ CM CM CM CM~ CM ~~~~~~~~

112

Li
_____ . .
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

0
(l
~

F-

!!ir
z
0
4::

t
~ ~1

CM — — — — — —
00
CM
4:’ OO
.
CM C
.
— CM CM CY~ CM C
— ——
0’. 0’. ~~‘N -‘ 0 0 5 0 0 0’.
CM
~~ —
—

— — CM CM — — — ‘ . i — — — i
-
. — CM ‘.
~~
— — —

113
..
- — ~~~~~ ~~~~~~~~~~~~~~~
. . . .-.——-
~~~~~~
-
 APPENDIX D

FACTOR ANALYSIS

OF

R OTTER QUE STIONS

114
...
.. .. . _ ..._ . .— . ,. .~
. . . ~~ -.—.——. — .- — .— - ---- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ — — — -.—..-. ~~~~~~~~~~~ . _ ,.... ~~.
.. .
,
 — ——
C CM CM CM C 00 ‘.0 ‘0

—
C’) CM ‘C N 00 0 —.. CM
2
•‘ CM —
c It) C
~~
— ‘~ ‘ — 0 0 CM 0’. — ~~ N ‘ .‘0 00 — C ‘C If) 00 00 ‘
CM C ) CM CM ~~ C — —0 C ~~
~~ ‘ C C — CM C C
~ ‘C 4:’ ~~
.
.
I’ I’
. • •
I I
. . •

‘.0
— CM 0’ CM N ~~ U) 0’. CM 0 0 0 0 0 CM 0’ 0 CM N C’) cc ~~ ‘ cc
2
CM
~~ — — C — C ‘~0 N — S N U) — 4:’ — 4:’ — CM
0’
U N C M — C — C — C~~~~~~ C N - - C C 4 ’: —
I’ I’ •
I

0
U)
U) — ‘C C’. If) 00 (N C’) N .‘0 C 0’ ‘C — .‘0 N ( NC M 0’ 0 0 0’
~ N
.
o i-~
U ) 0 ’ NC 4 :’ C 0’ ... N N 0’
0’ ~~
C U ’) 0 ’ \ O0 ’.0 C’) ‘.0 —

0’~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~ — LI,
I I I I I I I

F-

—
‘0 It) C U) — cc N — CM 0’

—
~ N ‘C 0’ ‘0 CM 0 0’ C
0
~~~~~ CC N N 0 0N 4 : ’ ’ C 0 ’ 4:~ ~~
C M L ’ ) 0 ’~~ C’) cc 0
~~ ‘
U
CM CM C — — ‘C C’) C ~ U) 0 — CM 0 ~~~~
0 — 0 ~~ C’) CM C -
.‘0
I I I I I I .... .

>..
1
-~
< CV)
Z C U) 0 ’ 4:~ 0 0 0 0 U ) U ) C M 0C
’0 0
C N c 0 CM oc ~~’) C ’ C
CM N

0
F-
~~
~
~~
. . •
l
•
I
•
~~~ I I I

CM

— — — —
1.4 C’) CM CM CM N 4:’ 00 CM 0’ It) I
I 00 0’ 00 U) 0’ CM C C C’) N U)
~ N C It) 00 CM 0 0 0 0 0 4:’
0 U) — N U) C’) — ‘.0 It) CM N C’) U) C’) 4:’ N CV’) —
0Cv ‘.0 CCu CM 5 — 0 — 0 — — ‘~~‘ C ‘C C ‘C
— U H
-~

,
—
~
0
CV) 4:’
— CV’) 0’ 0’. NO’ it) C
— — N C’))!I ‘.0 ‘0
4:’ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
—
~~ C C 0~~ S CM C ‘C CM — N C (NO
— C’) N CM CV) 0’

—
C’. It)

I I ,
~~ V
C)
1.~~~ ~~

(N U) ’ C5O~~~~~~~~~~~~~~~~~~~~
’ g
~
~~~
.ai
~
‘
~~
.
~~~ ~~~~

115
_ _ _ _ _ _ _ _ _
F

APPENDIX E

CORRELATION MATRIX

OF

LEADER REWARD BEHAVIOR QUESTIONS

116

I.
 _ _ _ _ _ _ _ _ _ _ _ _ ~~~~~~~~~ -- ~~~~~~~~~~~~ -~~~~-~~-- -—-—--
- -— ~~- ... __—.-—-~..

— CM C’) — U) .0 4:’ 0’. C’) S c c C C’ .U)

c c Oe ’—) C ’) 0 0 4:4 tt) 0 C CM N
~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
•
I I I —

U) C’) U) U) 0’ ~~ — (N C’- U) U) c c c c sU) 4:’ CM cc’ .0 4:’ ‘ C’) C N C

Z C’) — 00 C ’ C C’) N CM 0’
CM N U) ’C U) 4:’ C’) CV) N CM C
2 ~~~~~~~~~~~~~~~~~~~~~~~
F-
(I)
J
~~~~~~~~ I
I — —

~~ C~~~~~~ C ’ C~~~~~~
— ,
~~~
I —

— —
~P’ C C
c c N’ C C M
0’C — C U ) U ) 0 ”.-.. .- 0 0 C M ’ . 0,-
‘C’ .
~~ S C (N C CM CM ~~ 5 — 0 0 C~~~~
M — ~~~~
— CV) — 0 — C’) 4:’ CM CV)

0 ~ ‘_.
~

4:’ S N U) 0 C U) C’) U) 4:’ 0’ O N C ’ C’) C’) ’0 4:’ It)

0 U) C V ) C ’ U ) 0 0 C MC C ’ ) t t )0 0 U) 4:’
~~~ C M ’ . ~~~~ — O U ~~~~
’.
~~ ~~~~~~~~~~~~~~~~~~~~~
I — I

0
U ) C M C C ’ )C 0 ’ C U) L f) 0’ 4 : ’ C ’) C M f) C ’) % O C O
N ’ C tf) C’) C
_
U) 0 CM C C C U) U) — U) .0 C U) 4:”C CM -‘C
~:]
~~ ‘ C ’ ~
. —
I I — I I —
~~

00 CV’) — CM 00 It) CM 0’. U) C’4 U) cc

N U) ’ 0 _N 5 ’ C N
~~~ C~~~~~~
‘C C’) N CM —~~~
4:’ CV) C CV) ~~~ 4:” .O ~~~~~~ 4:~ C ‘.
~~
F- ~~~~
—
—

U) C’) C U) C U ) c V ) U ) . CM
~~ ~3 u
s-i I —
i-i
-

— —
11) 0 0~~ ’C ’) CV1 U’) S C C ’ ) ’
SC N C M N 0” CC 0’ 4:’ CMCC’)
0 0 CM 0 CM CM CM CM CM .-. If)
~~ 1 . U—;
i-i

—
~~
N CM CM
— ‘C .‘0 4:’ 0’. 0
C V ) C M C M 4 : C M.s- (N

117

~I. .
~~~~~~~~~ --
0

00
0
— U)
—4: ’

——
22
0’ C’
<0
1.4 1.4
0 0

. .0
~~~

F-
4::
Cac ~

bO bO
2 1.4 1.4
F- 0’ c c O C ’ .4:4 (N C M 0’. CV’) 0 c c0 ’5 0 0” U’) ~~~~~ ‘C C’)
(N
-
(N ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- 0 0
~

ii
Q) Q)
1.4 1.4
CM’C CM CM 0’ 0 ‘0 ( NO 4:’ 4:4 0’ U) 0’.C’) 00 0 0 ( N U ) 0 0
CM 0 00 ’0’ 4:’ 00 ( N 4 :4 ‘.0’0 U) CV) CM ‘0 ‘ ( N O (N.-
— ‘C 0
(N 4:’ C CV) C”) C 4:’ .-. 4:’ C~) 4:4 C’) CM 4:’ 4:’ 11)
~~
4:’ (NIt) 4:4 C’) ~~
I —
mm

N ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

118

~~~~~ -— .
~~
--
~~~ — .-
 _ _ _ _ _ _ _ _ _
_ _ _ _ _
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

APPENDIX F

CORRELATI ON MATRIX

FOR
REGRESSION ANALYSIS VARIAB LE S

119 ~~~
.
Iih
Il iIsIlIl.SIIJL _ ... _ __ _ _ . .
.
. — .-—--—— — —-——. —. . .
.
~ —.—..——.—-—.—.-
. . — .. . .—. ———— .— ~~~~~~~~~~~~~~~~~~~~~~ .-
,
_ _ _ _ _ _ _ _ _

C’) C
~~ c c ’ C ~~
— CM

I
~~~~I I I I I
I I I I
~~~~~

Cl)

— — z
Cl)
CM ’.0 0’. N ~~‘ CM — C’)
C’) F-

I I I~~~~~
I I • . . • • . ;
~~~

• I I I 0..
I I I I I I I I ~

4:: 5 U) ’C C M’0Q ~~ ‘ ~~ 0’ C’) C’) C’) (N 00 C’) C’. C’ 0’ (N

C ~~ ‘ C
(N C’) . ‘ — CM C”) ~~ C C’) 0 0 C’ ’ . C4:’ 0’~~~
C ~~
~ 0
~ ‘ ‘ C0 0 ~~
U) ‘ C C0 CC 0C O~~
• .‘~
Z I I I I I I
O
•
>... <

0)
‘. 0
-#5
5 1_a

O . . ‘ I I I I
• I I I I I

U)—
CC
-

22

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I
~~~ Z < ~~~~~~~~~~~~~~~~~~~~
J 4.l L~~~~~~~~~~~~~~~~~~ Cl) U) .-~~ -]
~
~~~
Cl) ~~~~~~

120
~ ~~~~~~~~~~~~~~~~~~~~~~~

U)
s-I
U)

C4CM
4:’ C C’
2 CMU)C

U)’.C~~~
(N
I

0000

-.3 _I

121

4
_4
.
-.
. - - -.—---— .
- .— .--—. -.-. —.-.— —. —- ~~~~~~~~~~~~~ - _.._.
. _ .—..-— _ — ~ .:‘
-‘fl—-- -— —..
- .-
~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - . --
 VIT A

Larry J. Corbin was born on 26 November 1949 in Columbus, Ohio.

He gradua ted fro m high school in Mt. Vernon , Ohio in 1967. He attended t

Ohio State Uni versity where he received a Bachelor of Science Degree in

Mathematics Education in 1971. At that ti me , he was commissioned a

Second Lieutenant in the United States Air Force. He was ini tially assigned

to the 1902 Communications Squadron , Hamilton AFB, California as squad-

ron Air Tra ffi c Con trol Officer. In 1973, he was reassigned to Kun san

AB , Korea as squadron Air Tra ffi c Control Officer and RAPC ON Chief

Controller. A subsequent assignment placed him as Chief, Air Tra ffi c

Control Operations, 1903 Communica tions Squadron. Devis-Mon tha n AFB,

Arizona in 1974 unti l ente ring the School of Engineering , Air Force

Institute of Technology, in 1976.

Cap tain Corbin is married to the forme r Susan Ybarra of La Cres-

enta, California. They ha ve one child; Cathy Lee.

F~ rmanent Address~ Rou te 1, Box 156A

- Lyndon , OH 45649

122

-- - —. - -~~~~ - -—.-
~~~~— --..-—
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
~~~~
_ _ - ~~~~~~~~~~~~~~~~~~ .
~~ ~~~. — - ~~~~~~~~~~ --~~~~~~~~~~

,
UNCLASSIFIED
S ECURITY CLASSI FICA TION OF T HIS RAGE ~~~~~~ Dale Enl.r.d)
READ WSTRUCT IONS
~ E~ ”~ ’
DA f ~~
~~ VV ~~
I W!J ..U
~ ~~ ~~~~~~~~
I~ I A I ~‘J~ ’ ~
3.
BEFORE COMPLET IN G FORM
R E CI P IE N T ’ S C A T A L O G NUMBER
I. REPORT NUMBER _. 2. GOVT ACCESSION NO

AFIT/GSM/SN/77S-2 ______________________
4. TITI.. E (end Sublsa.) I”- S. TYP E OF REPORT B PERIOD COV E R E G

PRODUCTIVITY AND JOB SATISFACTION IN

RESEARCH AND DEVELO P1V~ NT: ASSOCIATED MS Thesis
INDIVIDUAL AND SUPERVISORY VARIABLES 6. P E R F O R M I N G ORG. REPORT NUMBER

1. A IJT HOR(a) L C O N T R A C T OR G R A N T NUMBE R(S)

Larry J. Corbin , Capt , USAF -

S. PERFORMING O R G A N I Z A T I O N NAM E AND ADOR ESS tO . PROG RAM ELEMENT . P R O J E C T . T ASK

A R E A & W O R K UNIT NUM B ER S
Air Force Institute of Techno1ogy(AFIT/EN~
Wright-Patterson AFB , Ohio L4.5L~.33
II. C O NT R O L L I N G OF FICE N A M E AN D A OOR E3S 12. REPORT D A T E

Air Force Institute of Technolo~~r(AFIT/EN ) September 1977

Wright-Patterson AFB , Ohio Ll.5L1.33 t 3. NUMBER Q F PA G E S

_________________________________________________ 132
14. MONITORING AGENCY NAME a ADDRESS(I1 djI(.renl 1,00, CenlzolIi n4 OWe.) 3. S E C U R I T Y CLASS. (ol thIs report)

UNC LA SSIFIED
IS. . G ECLA SS IFICA TIO N DO W N G R A D I N G
SC.4ED UL E

!6. D I S TR I B U T I O N S T A T E M E N T (of thi. Report)

Appro’,ed for public release ; distributior’. ‘zL iit ed

~~

17. DISTRI B UTIO N STA T E M E N T (of A . abstrac t .nt.r.d in Block 20, ii dJl~.rsnt (soO, R.poa ~ )

14. S U P P L E M EN T A R Y NOTES
~Approved- ~b r—pub1ic release; lAW AFR 190-17
—
‘
-

,,
..
‘. ‘
- - . ~~~~~~ -
~~~~~~~
Jerral F. Gizess, Captain , IJSAF
~~~~~~~~~~~ ~~ ~~~~~~

Director of Information
S. CCV WO RDS (Continue on ,.v.,a. aid. I n.c. .e~~y end Id.nlify by block i,senb.r)
~
PRODUCTIVI TY
JOB SATISFACTION
RESEARCH AND DEVELOP !~~NT
SUPERV ISORY
20. A B S T RA C T (C.ntinu. on r...ve. aid. II n.c.. .vy id Identity by block ~.o
nbi,)

T he relationsh ips o f supervis o ry and individual variables t

the productivity and job satisfaction of scientists and engineers
in an AF R&D laboratory were explored. Productivity (quantity of
output ) and job satisfaction (Hoppock) were used as criterion .
The Supervisory Behavior Description Questionnaire , the Leader
Reward Behavior Instrument , and the Rotter Internal-External Scale
of Control of Reinforcement were used as predictors. Information
was obtained on 326 scientists/engineers. Factor analysis o f the
DD 1473 EDITION OF I NOV II ii OS$O I.ETI UNCLA SSIFIED
~~~ , ..
. Den. &,Ien .d) SE CURITY CLA UIFICATIOII OF THIS P& GI (~~
~ — ---. - - -. . . -

UNCLASSIFIED
S ECURITY CLASSIFICATION OF THIS PAGC(W I,en Date Ent., d)

20. Rotter scale indicates that the scale has questionable . ‘a-
1.idity , is multidimensonal, and may be sensitive to population dif-
~erences. For both the nonsupervisory scientists/engineers and the
group leaders , the following relationships were foun d. No relation-
ship was foun d between productivity and job satisfaction . Although
iigher e ducation , grade , and experienc e were associated with higher
productivity , no single predictor variable was shown to be si iif-
icantly associated with ail six of the pro ductivity variables, ~ in-
~luding leader behavior and the total Rotter score . However,
~orisider ation and positive leader reward behavior were positively
related to job satisfaction. Education was found to be associated
~o job sat isfaction : po sitively for the nonsupervisory scientists/
?ngineers and negatively for the group leaders . The total Rotter
score was negatively associated with job satisfaction for the non-
supervisory scientists/engineers.

IJNC LASS IFIED

SECURITY C L A S S I F I C A T I O N OF THIS PA GE(WP ,en 0.t. Ent .red )

- -
.— ._ ~~: u ~~~~~~~~~~~ —~~~~~~-~--~~ -u~~— _ —-~~~~~~4