Motivation Through The Design of Work: Test of A Theory

ORGANI ZATI ONAL BEHAVI OR AND HUMAN PERFORMANCE 1 6 , 2 5 0 - 2 7 9 (1976)
Motivation through the Design of Work: Test of a

Theory
] . RI CHARD HACKMAN
Yale University
AND
GREC R. OLDHAM
University of lllinois
A model is pr oposed t hat specifies t he condi t i ons under whi ch i ndi vi dual s
will become i nt ernal l y mot i vat ed t o per f or m effect i vel y on t hei r j obs. The
model f ocuses on t he i nt er act i on among t hr ee cl asses of vari abl es: (a) t he
psychol ogi cal st at es of empl oyees t hat must be pr esent for i nt ernal l y mot i -
vat ed wor k behavi or t o devel op; (b) t he char act er i st i cs of j obs t hat can creat e
t hese psychol ogi cal st at es; and (c) t he at t r i but es of i ndi vi dual s t hat det er mi ne
how posi t i vel y a per son will r espond t o a compl ex and chal l engi ng j ob. The
model was t est ed for 658 empl oyees who wor k on 62 di fferent j obs in seven
organi zat i ons, and r esul t s suppor t i t s validity. A number of speci al feat ures of
t he model are di scussed (including its use as a basi s for t he di agnosi s of j obs
and t he eval uat i on of j ob redesi gn proj ect s), and t he model is compar ed t o
ot her t heor i es of j ob desi gn.
Work redesi gn is becomi ng increasingly promi nent as a strategy for
attempting t o i mprove si mul taneousl y the producti vi ty and the quality o f
the work experi ence o f empl oyees in cont emporary organi zati ons. Al-
though t he benefi ts of work redesi gn (or "job enri chment" or "job en-
l argement") are wi del y t out ed in the management literature, in fact little is
known about the reasons why "enri ched" work somet i mes l eads to posi-
ti ve out c ome s for workers and for their empl oyi ng organi zati ons. Even
l ess is known about the rel ati ve ef f ect i venes s o f vari ous strategies for
carrying out the redesi gn o f work ( Hackman, 1975).
One reason for this state o f affairs is that exi sti ng theori es o f work
The aut hor s expr ess gr eat appreci at i on to member s of t he consul t i ng fi rm t hat hel ped us
gain access t o t he organi zat i ons wher e t hi s r es ear ch was conduct ed; t o Kennet h Br ousseau,
Dani el Fel dman, and Li nda Fr ank for assi st ance in dat a col l ect i on and anal ysi s; and to
Gerri t Wol f for hel p in anal yt i c pl anni ng.
Thi s r epor t was pr epar ed i n connect i on wi t h r es ear ch suppor t ed by t he Office of Naval
Resear ch (Organi zat i onal Ef f ect i veness Resear ch Program, Cont r act No. N00014-67A-
0097-0026, NR 170-744), and by t he Manpower Admi ni st r at i on, U. S. Depar t ment of Labor
( Resear ch and Devel opment Gr ant No. 21-09-74-14). Si nce grant ees conduct i ng r es ear ch
under gover nment sponsor shi p are encour aged t o expr ess t hei r own j udgment freely, t hi s
r epor t does not necessar i l y r epr es ent t he official opi ni on or policy of t he gover nment .
250
Copyright 1976 by Academic Press, Inc.
All rights of reproduction in any form reserved.
MOT I VAT I ON T HR OUGH DE S I GN OF WOR K 251
redesign are not fully adequat e to meet the probl ems encount er ed in t hei r
application. Especially t roubl esome is t he pauci t y of concept ual tools that
are di rect l y useful in guiding the implementation and evaluation of work
redesign projects. In the paragraphs to follow, we exami ne several exist-
ing t heoret i cal approaches to wor k redesign, wi t h a special eye t oward t he
measurability of t he concept s empl oyed and t he action implications of t he
theorizing (cf. Porter, Lawler, & Hackman, 1975, Chap. 10). We t hen
propose and report a test of a t heor y of work redesi gn that focuses spe-
cifically on how the characteristics of jobs and the characteristics of people
interact to det ermi ne when an " enr i ched" job will lead to beneficial out-
comes, and when it will not.
Theoretical Approaches to Work Redesign
Motivation-hygiene theory. By far t he most influential t heor y rel evant
to wor k redesign has been t he Herzberg t wo-fact or t heory of satisfaction
and motivation (Herzberg, Mausner, & Snyder man, 1959; Herzberg,
1966). In essence, t he t heor y proposes that t he pri mary det ermi nant s of
empl oyee satisfaction are factors intrinsic to the work that is done (i.e.,
recognition, achi evement , responsibility, advancement , personal growt h
in compet ence). These factors are called "mot i vat or s" because t hey are
believed to be effective in motivating empl oyees to superior effort and
performance. Dissatisfaction, on t he ot her hand, is seen as being caused
by "hygi ene f act or s" that are extrinsic to the work itself. Exampl es in-
clude company policies, supervi sory practices, pay plans, working condi-
tions, and so on. The Herzberg t heor y specifies that a j ob will enhance
work motivation and satisfaction onl y to t he degree that "mot i vat or s" are
designed into the wor k itself. Changes that deal solely with " hygi ene"
factors should not lead to i ncreases in empl oyee motivation.
It is to t he credi t of t he Herzberg t heor y that it has prompt ed a great
deal of research, and inspired several successful change projects involv-
ing the redesign of work (e.g., Ford, 1969; Paul, Robert son, & Herzberg,
1969). Yet t here are difficulties with the t heor y that to some ext ent com-
promise its usefulness.
For one, a number of researchers have been unable to provide empirical
support for the major t enet s of t he t wo-fact or t heor y itself (see, for exam-
ple, Dunnet t e, Campbell, & Hakel, 1967; Hinton, 1968; King, 1970. For
anal yses f avor abl e to t he t heor y, see Her zber g, 1966; Whi t set t &
Winslow, 1967). It appears that the original di chot omi zat i on of aspects of
the work-place into "mot i vat or s " and "hygi ene f act or s" may have been
largely a funct i on of methodological artifact, and t he present concept ual
status of t he t heor y must be consi dered highly uncertain.
Moreover, t he t heor y does not provide for di fferences among people in
how responsi ve t hey are likely to be to " enr i ched" jobs. In t he AT&T
252 HAC KMAN AND OL DHAM
studies based on t he t heor y (Ford, 1969), for exampl e, it was assumed that
t he motivating fact ors potentially coul d i ncrease t he wor k motivation of
all empl oyees. Yet it now appears that some individuals are much more
likely to respond positively to an enri ched, compl ex j ob t han are ot hers
(Hulin, 1971). The t heor y provides no help in det ermi ni ng how such indi-
vidual di fferences phenomena should be dealt with, ei t her at t he concep-
tual level or in t he actual applications.
Finally, t he t heor y in its present form does not specify how t he pre-
sence or absence of motivating factors can be measured for existing jobs.
At t he least, this i ncreases t he difficulty of testing t he t heory in on-going
organizations. It also limits t he degree to whi ch t he t heor y can be used to
diagnose j obs prior to pl anned change, or to eval uat e t he effects of wor k
redesign activities aft er changes have been carri ed out.
Activation theory. While psychologists have for many years studied t he
ant ecedent s and consequences of hei ght ened and depressed levels of
psychological and physiological activation in organisms (Berlyne, 1967),
only recent l y have at t empt s been made to use activation t heor y to under-
stand t he work behavi or of individuals in organizations. Scott (1966) has
r evi ewed a number of studies that show how people react to chroni c
states of underact i vat i on at wor k by engaging in arousal-enhancing be-
haviors, some of whi ch have clearly dysfunct i onal consequences for wor k
effect i veness. The findings Scot t summari zes suggest t hat act i vat i on
t heor y may be of consi derabl e use in underst andi ng j obs that are highly
r epet i t i ve- - and in planning for t ask designs that minimize t he dysfunc-
tional consequences of underact i vat i ng work. Activation theorists have
given rel at i vel y little at t ent i on to j obs t hat may be overst i mul at i ng,
perhaps because few such j obs exist for rank-and-fi e workers in cont em-
porary organizations.
While activation t heor y clearly has consi derabl e rel evance to bot h t he
t heor y and pract i ce of j ob design, t wo t hor ny probl ems must be dealt with
before t he t heor y can be fully applied to real-world j ob design problems.
First, means must be devel oped for measuri ng current levels of activation
of individuals in actual wor k settings (cf. Thayer, 1967), and for assessing
t he "opt i mal l evel " of activation for di fferent individuals. Until such
methodologies are devel oped, it will remai n impractical to use activation
t heor y in predicting or changing empl oyee react i ons to t hei r j obs except in
a ver y gross fashi on; e. g. , in situations wher e it is cl ear t hat most
empl oyees are enormousl y over- or underst i mul at ed by their jobs.
A second probl em has to do with ambiguities regarding t he processes
by whi ch individuals adapt to changing levels in stimulation. Individuals'
levels of activation decr ease markedl y as a funct i on of familiarity with a
given stimulus situation. However , after a peri od of rest , re-present at i on
of t he same stimulus situation will once again raise the level of activation
(Scott, 1966). More complete understanding of the waxing and waning of
activation in various circumstances could have many implications for job
design practices; for example, the practice of "job rotation." Those who
advocate job rotation claim that work motivation can be kept reasonably
high by rotating individuals through several different jobs, even though
each of the jobs would become monotonous and boring if one were to
remain on it for a long period of time. If future research can identify ways
to maintain activation at near-optimal levels through planned stimulus
change, then the theory can contribute substantially to increasing the
usefulness of job rotation as a motivational technique. If, however, it
turns out that there are general and inevitable decreases in activation over
time regardless of how different tasks and rest periods are cycled, then the
long-term usefulness of the technique would seem to be limited.
In either case, the potential for applying activation theory to the design
of jobs may be limited mainly to those cases in which there are actively
dysfunctional affective and behavioral outcomes associated with routine,
repetitive jobs. The theory offers less guidance for the design of work that
will elicit and maintain positive and self-reinforcing work motivation.
Socio-technical syst ems theory. The socio-technical systems approach
to work redesign provides significant insight into the interdependencies
between technical aspects of the work itself and the broader social milieu
in which the work is done (Emery & Trist, 1969; Trist, Higgin, Murray, &
Pollock, 1963). The theory has evolved from (and has been used as an
explanatory device for) numerous planned changes of work systems.
Many of these experiments have provided vivid illustration of the interac-
tions between the social and technical aspects of the workplace, and at the
same time have proven successful as action projects--in that beneficial
outcomes were obtained both for employees and for the organizations in
which they worked (cf. Davis & Trist, Note 1; Rice, 1958). Of special
interest is the contribution of socio-technical systems theory in develop-
ing the notion of the "autonomous work group," in which members of a
work team share among themselves much of the decision-making having
to do with the planning and execution of the work (Gulowsen, 1972;
Herbst, 1962). Creation of autonomous work groups promises to become
increasingly prominent and useful as a strategy for redesigning and im-
proving work systems.
Yet for all its merit, the socio-technical systems approach provides few
explicit specifications of how (and under what circumstances) the work
itself and the social surroundings affect one another. It is, therefore,
difficult to test the adequacy of the theory qua theory. Moreover, the
approach provides little specific guidance about how (and how not to)
proceed in carrying out work redesign activities, other than the general
dictum to attend to both the technical and social aspects of the work
254 HACKMAN AND OLDHAM
setting and the devi ce of the aut onomous wor k group. Absent from the
approach, for exampl e, are explicit means for diagnosing a wor k syst em
prior to change (to ascert ai n what " s houl d" be changed, and how), or for
evaluating in syst emat i c t erms the out comes of changes that have been
compl et ed.
For t hese reasons, t he major value of socio-technical syst ems t heory
appears to be its consi derabl e useful ness as a way of thinking about wor k
syst ems and their redesign. In its present form, it has onl y limited use in
generat i ng new under st andi ng t hr ough quant i t at i ve t est s of t heor y-
specified proposi t i ons, or in providing explicit and concr et e guidance
about what organizational changes to make under what ci rcumst ances.
Jobs and individual differences: An interactive approach. Resear ch on
wor k design that f ocuses on the obj ect i ve charact eri st i cs of j obs is r oot ed
in t he wor k of Tur ner and Lawr ence (1965). These r esear cher s de-
vel oped measures of six "Requi si t e Task At t r i but es" that wer e predi ct ed
to relate positively t o empl oyee satisfaction and at t endance. A summary
measure, the Requi si t e Task At t ri but es Index (RTA Index) was deri ved
from the six measures and used to t est relationships bet ween the nat ure of
j obs and empl oyee react i ons to them.
Expect ed positive relationships bet ween t he RTA Index and empl oyee
satisfaction and at t endance were found only for wor ker s from fact ori es
l ocat ed in small t owns. For empl oyees in urban wor k settings, satisfaction
was i nversel y rel at ed to the scores of j obs on the RTA Index, and absen-
t eei sm was unrel at ed to the Index. The investigators concl uded that reac-
tions t o j obs high on the RTA Index wer e moder at ed by di fferences in t he
cultural backgrounds of empl oyees. Subsequent r esear ch by Bl ood and
Hulin (Bl ood & Hulin, 1967; Hulin & Bl ood, 1968) provi des support for
t he not i on that subcul t ural fact ors moder at e wor ker responses to the de-
sign of their j obs.
A st udy by Hackman and Lawl er (1971) provi des further evi dence that
j ob charact eri st i cs can directly affect empl oyee attitudes and behavi or at
work. These aut hors suggest ed that empl oyees should react posi t i vel y to
four " c o r e " dimensions adapt ed from t hose used previ ousl y by Turner
and Lawr ence (i.e., vari et y, t ask identity, aut onomy, feedback). In addi-
tion, Hackman and Lawl er pr oposed that individuals who were desi rous
of growt h satisfactions at wor k should r espond especially positively to
j obs high on t he core dimensions, since t hese individuals are most likely
to val ue the kinds of opport uni t i es and internal rewards that compl ex
j obs offer.
Resul t s of the st udy generally support ed the hypot hesi s t hat empl oyees
who wor k on j obs high on the core dimensions show high wor k motiva-
tion, sat i sfact i on, performance, and at t endance. Also, Hackman and
Lawl er found that a number of dependent measures were moderat ed as
predi ct ed by growt h need strength: That is, empl oyees with high meas-
ured needs for growt h r esponded more posi t i vel y to compl ex j obs than
did empl oyees l ow in growt h need strength.
The appropri at e concept ual i zat i on and measur ement of the di fferences
among peopl e that moderat e how t hey respond to compl ex j obs has been
the subj ect of a number of recent studies. Findings similar to t hose re-
port ed by Hackman and La wl e r have been r epor t ed by Bri ef and Aldag
(1975), by Ol dham (in press), and by Sims and Szilagyi (Not e 4), using a
measure of growt h need strength (although the Bri ef and Aldag st udy
provi ded onl y partial replication). Support i ve findings also have been ob-
tained by Robey (1974), using as an individual difference measure "ext ri n-
si c" vs "i nt ri nsi c" wor k values. Failures to obt ai n a moderating effect
have been report ed by Shepard (1970) (using a measure of alienation from
work) and by St one (1976) (using a measure of empl oyee endor sement of
the Prot est ant wor k ethic). Wanous (1974) directly compar ed the useful-
ness of (a) higher order need strength, (b) endor sement of the Prot est ant
wor k ethic, and (c) urban vs rural subcul t ural background as moderat ors
of j ob effects. All t hree variables were found to be of some value as
moderat ors, with the need strength measure st rongest and the urban
- rural measure weakest .
In sum, there is now substantial evi dence that di fferences among peopl e
do moderat e how t hey react to the compl exi t y and challenge of their
work, and studies using direct measur es of individual needs seem to pro-
vide more consi st ent and strong support for this finding than do measur es
of subcultural background or of generalized wor k values.
THE JOB CHARACTERISTICS MODEL
The model present ed and t est ed in this paper is an at t empt to ext end,
refine, and syst emat i ze the relationships descri bed above bet ween j ob
charact eri st i cs and individual r esponses to t he work. The basi c j ob
charact eri st i cs model is present ed in Fig. 1. At the most general level, five
" c o r e " j ob dimensions are seen as prompt i ng three psychol ogi cal states
which, in turn, lead to a number of beneficial personal and wor k out-
comes. The links bet ween the j ob di mensi ons and the psychol ogi cal
states, and bet ween the psychol ogi cal states and the out comes, are shown
as moder at ed by individual growt h need strength. Each of the major
classes of variables in the model is di scussed in more detail bel ow.
Psychological States
The t hree psychol ogi cal st at es (experi enced meani ngful ness of the
work, experi enced responsibility for the out comes of the work and knowl-
edge of t he results of t he wor k activities) are t he causal core of t he model.
Following Hackman and Lawl er (1971), the model post ul at es that an indi-
vidual experi ences positive affect to the ext ent that he learns (knowledge
I O O R E J o s I J 1
L PSYCHOLOGICAL
DIMENSIONS I ~ l S T A T E S
~ L Experienced ~]
Skill Variety
Task I d e n t i t y f . ~- Meaningfulness ]
| of the Work |
Tank S ig n if ic a n c e . , , J /
/
Experienced L ~
, u , . . . . , , " " : 'r r
of the Work |
|
Knowledge f the 1
. A o l % R e o , s O f . w r k A c l . v t , e |
~ , { EMPLOYEE GROWTH~%~ ~
~NEED S T R E N G T H J
FIG. 1. The j ob characteri sti cs mo de l o f wo r k mot i vat i on.
P E R S O N A L AND I
>
l
WORK OUTCOMES
High Internal
Work Motivation
High Quality
Work Performance
High Satisfaction
With 1he W ork,
Low Absenteeism
and Turnover
o f resul t s) that he personal l y ( experi enced responsi bi l i t y) has perf ormed
wel l on a t ask that he cares about ( experi enced meani ngf ul ness) .
Thi s pos i t i ve af f ect is rei nforci ng t o t he i ndi vi dual , and s erves as an
i ncent i ve for hi m t o cont i nue t o try t o perf orm wel l in t he future. Whe n he
d o e s n o t p e r f o r m we l l , he d o e s n o t e x p e r i e n c e an i nt e r na l -
l y rei nforci ng state o f affairs, and he may el ect t o try harder in t he future
s o as t o regai n t he internal rewards that g o o d perf ormance bri ngs. The net
resul t is a sel f - perpet uat i ng c yc l e o f pos i t i ve wor k mot i vat i on powe r e d by
sel f - generat ed rewards, that i s predi ct ed t o cont i nue unti l one or more o f
t he three ps yc hol ogi c al st at es i s no l onger present , or unti l t he i ndi vi dual
no l onger val ues t he internal rewards that deri ve f rom g o o d perf ormance.
It s houl d be not e d that sel f - generat ed mot i vat i on s houl d be hi ghest
whe n all three o f the ps yc hol ogi c al states are present. I f t he performer
f eel s ful l y res pons i bl e f or wor k o ut c o me s on a meani ngf ul task, but never
finds out ho w we l l he i s perf ormi ng, it i s doubt f ul that he wi l l experi ence
t he internal rewards that can prompt sel f - generat ed mot i vat i on. Si mi l arl y,
i f he has full knowl e dge o f t he resul t s o f t he work, but experi ences t he
t ask as trivial (or f eel s no personal responsi bi l i t y for t he resul t s o f t he
work) , internal mot i vat i on will not be high.
The three ps yc hol ogi c al st at es are def i ned as f ol l ows :
Expe r i e nc e d Meani ngf ul nes s o f t he Work. The degree to whi c h t he i ndi vi dual ex-
peri ences t he j o b as one wh i c h i s general l y meani ngf ul , val uabl e, and wor t hwhi l e ;
Expe r i e nc e d Re s pons i bi l i t y for Work Out c ome s . The degree to wh i c h t he i ndi -
vi dual f e e l s pers onal l y account abl e and responsi bl e for t he resul ts o f t he wor k he or
s he doe s ;
Knowledge of Results. The degree to which the individual knows and understands,
on a continuous basis, how effectively he or she is performing the job.
Job Dimensions
Of t he five char act er i s t i cs o f j obs s hown in Fig. 1 as f ost er i ng t he
e me r ge nc e of t he ps ychol ogi cal st at es, t hr ee cont r i but e t o t he exper i -
e nc e d meani ngf ul ness of t he wor k, and one each cont r i but es t o exper i -
e nc e d r esponsi bi l i t y and t o knowl edge of r esul t s.
Toward experienced meaningfulness, Thr e e j ob char act er i s t i cs com-
bi ne addi t i vel y t o det er mi ne t he ps ychol ogi cal meani ngf ul ness of a j ob.
Th e y are:
(1) Skill Variety. The degree to which a job requires a variety of different activities
in carrying out the work, which involve the use of a number of different skills and
talents of the person.
Whe n a t ask r equi r es a pe r s on t o engage in act i vi t i es t hat chal l enge or
s t r et ch his skills and abi l i t i es, t hat t ask al most i nvar i abl y is e xpe r i e nc e d as
meani ngf ul by t he i ndi vi dual . Ma ny par l or games, puzzl es, and r ecr ea-
t i onal act i vi t i es, f or exampl e, achi eve muc h of t hei r f asci nat i on be c a us e
t he y t ap and t est t he i nt el l ect i ve or mot or skills of t he peopl e who do
t hem. Whe n a j ob dr aws upon s ever al skills of an e mpl oye e , t hat indi-
vi dual ma y find t he j ob t o be of e nor mous per s onal meani ng- - - even if, in
any absol ut e sense, it is not o f gr eat si gni fi cance or i mpor t ance.
(2) Task Identity. The degree to which the job requires completion of a "whole" and
identifiable piece of work; that is, doing a job from beginning to end with a visible
outcome.
If, f or exampl e, an e mpl oye e as s embl es a c ompl e t e pr oduc t ( or pr o-
vi des a compl et e uni t of ser vi ce) he shoul d find t he wor k mor e meani ngf ul
t han woul d be t he cas e i f he wer e r es pons i bl e f or onl y a smal l par t of t he
whol e j ob, ot he r t hi ngs ( such as skill var i et y) a s s ume d equal .
(3) Task Significance. The degree to which the job has a substantial impact on the
lives or work of other people, whether in the immediate organization or in the external
environment.
Whe n an i ndi vi dual under s t ands t hat t he r esul t s of his wor k ma y have a
si gni fi cant ef f ect on t he wel l - bei ng o f ot he r peopl e, t he meani ngf ul ness of
t hat wor k usual l y is enhanced. Emp l o y e e s who t i ght en nut s on ai r cr af t
br ake assembl i es, f or exampl e, ar e muc h mor e l i kel y t o per cei ve t hei r
wor k as meani ngf ul t han ar e wor ker s who fill smal l boxes wi t h paper
cl i ps - - agai n, e ve n t hough t he skill l evel s i nvol ved ma y be compar abl e.
Toward experienced responsibility. The j ob char act er i s t i c pr edi ct ed t o
pr ompt e mpl oye e feel i ngs of per s onal r esponsi bi l i t y f or t he wor k out -
c ome s is a ut onomy. To t he ext ent t hat a j ob has hi gh a ut onomy, t he
out comes depend increasingly on the individual' s own efforts, initiatives,
and deci si ons rat her t han on the adequacy of instructions from t he boss or
on a manual of j ob procedures. In such ci rcumst ances, the individual
shoul d feel strong personal responsibility for the success and failures that
occur on the j ob. Aut onomy is defined as follows:
Autonomy. The degree to which the j ob provides substantial freedom, independence,
and discretion to the individual in scheduling t he work and in determining the proce-
dures to be used in carrying it out.
Toward knowledge of results. The j ob characteristic that fosters knowl-
edge of results is f eedback, whi ch is defined as follows:
Feedback. The degree to which carrying out the work activities required by the j ob
results in the individual obtaining direct and clear information about the effectiveness of
his or her performance.
Summary: The overall "motivating potential" of a j ob. Accordi ng to
the j ob charact eri st i cs model, the overall potential of a j ob to pr ompt
internal wor k mot i vat i on on the part of j ob i ncumbent s should be highest
when all of the following are true: (a) the j ob is high on at least one (and hope-
fully more) of the three j ob dimensions that lead to experi enced meaningful-
ness, (19) the j ob is high on aut onomy, and (c) the j ob is high on feedback.
The Motivating Potential Score (MPS) is a measure of the degree to
whi ch t he above conditions are met. MPS is comput ed by combining the
scores of j obs on t he five di mensi ons as follows:
Mot i vat i ng= r Skill + Task + Task q
Potential LVariety Ident i t y Significance..J x Aut onomy x Feedback
Score (MPS) - 3
As can be seen from the formula, a near-zero score of a j ob on either
aut onomy or f eedback will reduce the overall MPS to near-zero; wher eas
a near-zero score on one of the t hree j ob dimensions that cont ri but e to
experi enced meaningfulness cannot, by itself, do so.
Individual Growth Need Strength
As not ed earlier, t here is now substantial evi dence that di fferences
among peopl e moderat e how t hey react to their work, and individual need
strength appears t o be a useful way to concept ual i ze and measure such
differences. The basic prediction is that peopl e who have high need for
personal growt h and devel opment will respond more positively to a j ob
high in motivating potential than peopl e wi t h low growt h need strength.
There are t wo possi bl e " s i t e s " for this moderat i ng effect in the motiva-
tional sequence shown in Fig. 1: (a) at the link bet ween the obj ect i ve j ob
dimensions and t he psychol ogi cal st at es, and (b) at the link bet ween the
psychol ogi cal st at es and the out come variables. The former woul d imply
that high growt h need peopl e are more likely (or bet t er able) to experience
the psychol ogi cal st at es when t he obj ect i ve j ob is good t han are their l ow
growt h need count erpart s. The latter allows the possibility that nearl y
ever ybody may experi ence the psychol ogi cal st at es when j ob conditions
are right, but that individuals with high growt h needs r espond more posi-
tively to that experi ence. It may be, of course, that growt h need strength
moderat es at both points in the sequence, as t ent at i vel y shown in Fig. 1.
Empirical t est s of t hese alternative "l ocat i ons " for the moderating effect
of growt h-need strength are r epor t ed later in this paper.
Outcome Variables
Al so shown in Fig. l are several out come variables that are predi ct ed to
be af f ect ed by t he l evel of j o b - b a s e d mot i vat i on e xpe r i e nc e d by
peopl e at work. Especi al l y critical to the t heor y is the measure of internal
wor k motivation (Lawl er & Hall, 1970; Hackman & Lawl er, 1971) be-
cause it t aps directly the cont i ngency bet ween effect i ve performance and
self-administered affective rewards. Typi cal quest i onnai re items measur-
ing internal wor k motivation include: (a) I feel a great sense of personal
satisfaction when I do this j ob well; (b) I feel bad and unhappy when I
di scover that I have performed poorl y on this j ob; and (c) My own feelings
are not affect ed much one way or the ot her by how well I do on this j ob
(reversed scoring).
Ot her out comes listed in Fig. 1 are the quality of wor k performance, j ob
satisfaction (especially satisfaction wi t h opport uni t i es for personal growt h
and devel opment on the j ob), absent eei sm, and t urnover. All of t hese
out comes are expect ed to be more positive for j obs wi t h high motivating
potential than for j obs l ow in MPS. Causal priorities among the several
out come variables are not explicitly addressed by t he model (cf. Oldham,
in press).
METHOD
The j ob charact eri st i cs model was t est ed using dat a obt ai ned from 658
empl oyees working on 62 different j obs in seven organizations. The j obs
are highly het erogeneous, including bl ue collar, whi t e collar, and profes-
sional work. Bot h industrial and servi ce organizations are included in the
sample, but all are busi ness organizations. The organizations are l ocat ed
in the East , Sout heast , and Mi dwest , in bot h urban and rural settings.
The pri mary dat a collection i nst rument was the Job Diagnostic Sur vey
(JDS), an i nst rument specifically designed t o measur e each of the vari-
ables in the j ob charact eri st i cs model. Propert i es of the JDS (including
descri pt i ons of item cont ent and format, and reliabilities of each measure)
are descri bed el sewhere ( Hackman & Oldham, Not e 3; 1975). Incl uded
t here are the means, st andard devi at i ons, and intercorrelations of JDS
measures for the r espondent s whose dat a are used in this report .
All dat a were col l ect ed on site by one of the aut hors or their associ at es.
One to four days wer e spent by the researchers at each organization
collecting data. Procedural st eps t ypi cal l y wer e as follows:
(1) The nat ure of t he r esear ch was expl ai ned to second- or third-level
management , and permi ssi on t o administer the i nst rument was secured.
Managers wer e i nformed t hat t he proj ect had to do wi t h t he refinement of
an i nst rument t o di agnose j obs, and that it woul d i nvol ve col l ect i on of
dat a from empl oyees, from their supervi sors, and from company records.
(2) The JDS was admi ni st ered to groups of empl oyees (ranging from 3
to 25 at a time). Before taking the quest i onnai re, empl oyees were told
about t he nat ure and pur poses of the r esear ch and wer e given t he opt i on
of not participating. Few empl oyees declined t o compl et e the question-
naire. It also was emphasi zed that all information obt ai ned woul d be held
in confidence, and t hat no one in t he organization woul d have access to
individual responses. Empl oyees were told t hat it was desi rabl e t o have
names on quest i onnai res for r esear ch purposes, but that this also was
vol unt ary. About 10% of the respondent s decl i ned t o provi de their names.
(3) Supervi sors and the researchers compl et ed the Job Rating For m
( Hackman & Oldham, Not e 3), whi ch measur es the charact eri st i cs of the
focal j ob as vi ewed by individuals who do not wor k on that j ob. Prior to
compl et i ng the Job Rating Form, t he researchers obser ved each j ob for
bet ween 1 and 2 hr.
(4) Member s of management wer e asked to rate t he wor k performance
of each respondent on (a) effort expended on the j ob, (b) wor k quality,
and (c) wor k quantity. The ratings were made on seven-poi nt scales de-
vel oped specifically for r esear ch purposes. Because the intercorrelations
among the three rating scales were high (median = .53), a summary meas-
ure of wor k effect i veness was obt ai ned by averaging ratings across the
t hree scales and across the supervi sors who rat ed each empl oyee. Only
t he summar y measur e is used in t he anal yses r epor t ed in this paper.
(5) Absence dat a wer e obt ai ned from company records. These dat a
wer e r ecor ded in t erms of the number of days each empl oyee in the
sampl e had been absent during the i mmedi at el y precedi ng year.
Empl oyee descri pt i ons of t he obj ect i ve charact eri st i cs of their j obs (us-
ing t he JDS) wer e compar ed t o similar descri pt i ons made by researcher-
obser ver s using the Job Rating Form. Medi an correl at i on bet ween the j ob
i ncumbent s and t he obs e r ve r s f or t he fi ve cor e di mensi ons is .65
( Hackman & Oldham, 1975). Evi dent l y empl oyees wer e able to provi de
rat her accurat e descri pt i ons of the charact eri st i cs of their j obs; hence,
MOTIVATION THROUGH DESIGN OF WORK 261
empl oyee-generat ed measur es are used for t est s of t he j ob charact eri st i cs
model. 1
For some j obs in some organizations it was not possi bl e to obt ai n com-
plete dat a for all variables. Therefore, some of t he resul t s r epor t ed in this
paper are based on that subset of the total sample for whi ch compl et e dat a
were available for the variable (or variables) of interest. Also, absence
reporting procedures and internal performance standards varied among
the seven organizations. Therefore, anal yses were performed separat el y
for each of the organizations whenever feasible, and medi an results are
report ed. In such cases, statistical significance was det ermi ned by com-
bining the p values obt ai ned in the seven separat e anal yses, following
procedures devel oped by St ouffer et al. (1949) and descri bed in Most el l er
and Bush (1954, p. 329).
RESULTS
The j ob charact eri st i cs model is sufficiently compl ex that it cannot be
t est ed in a single analytic step. Therefore, t hree separat e groups of
analyses are r epor t ed bel ow, each of whi ch bears on a different aspect of
the model.
(1) Simple analysis (by zero-order correlation) of the relationships of
the j ob dimensions and the psychol ogi cal st at es with the out come vari-
ables.
(2) Analysis (by partial correl at i on and multiple regression) of the de-
gree t o whi ch t he psychol ogi cal st at es medi at e bet ween j ob characteris-
tics and out come variables as predi ct ed.
(3) Test of the degree to whi ch empl oyees' react i ons to their wor k are
moderat ed by individual growt h-need strength as specified by the model.
Relationships of the Job Dimensions and Psychological States with
the Outcomes
The medi an correlations of the j ob dimensions and the psychol ogi cal
states with each out come measur e are shown in Tabl e 1. In general,
resul t s are consi st ent wi t h expect at i ons from t he model: Correl at i ons are
in the predi ct ed direction, and most achi eve accept abl e levels of statistical
significance. The psychol ogi cal st at es (whi ch in t he model are im-
medi at el y causal of the out comes) generally correl at e higher with the
1 It can reasonably be argued that when the intent is to predict or understand employee
attitudes or behavior at work (as is presently the case), employee ratings of the j ob dimen-
sions are preferable to use, since it is an employee' s own perceptions of the objective job
that is causal of his reactions to it (cf. Hackman & Lawler, 1971).
out come measures than do the j ob dimensions. The summary Motivating
Potential Score (MPS) relates more strongly to the out comes than do any
of its component j ob di mensi ons, also as expect ed. Relationships involv-
ing absent eei sm and performance, however , are not as strong as expect ed
and are generally smaller than relationships involving the measures of
satisfaction and mot i vat i on (perhaps simply because absent eei sm and per-
formance do not share common met hod variance wi t h the j ob characteris-
tics and the psychol ogi cal states).
Test of the Mediating Function of the Psychological States
The j ob charact eri st i cs model specifies that t he t hree psychol ogi cal
st at es medi at e bet ween the j ob charact eri st i cs and the out come variables
(Fig. 1). The validity of this general proposi t i on is t est ed by asking t hree
research questions. First, are predi ct i ons of the out come measures from
the psychol ogi cal states maxi mi zed when all t hree of the psychol ogi cal
st at es are used, or are equally strong relationships obt ai ned using the
psychol ogi cal states singly or in pai rs? Second, are the relationships be-
t ween the j ob dimensions and t he out come measures empirically depen-
dent on the psychol ogi cal states, or do the j ob dimensions predi ct the
out come measur es j ust as well if the psychol ogi cal states are ignored?
And third, do specific j ob dimensions relate to specific psychol ogi cal
st at es as specified in the model , or are the t wo sets of variables related
more compl exl y (or less so) t han predi ct ed?
These t hree quest i ons are addressed separat el y bel ow. To maximize the
stability of t he results (many of whi ch are based on partial and multiple
correlations), all 658 subj ect s were used in the anal yses for the three
out come vari abl es t hat wer e measured identically in all organizations in
the sample: internal motivation, general satisfaction, and growt h satisfac-
tion.
Are all three psychological states necessary to maximize prediction of
the outcome measures? To t est this quest i on, regressi ons wer e comput ed
predicting the out come measures (a) from each of t he t hree psychol ogi cal
st at es t aken alone (i.e., the zero-order correlations), (b) from the three
possi bl e pairs of t he psychol ogi cal st at es, and ( c) f r om all t hree psychol og-
ical st at es t aken together. Resul t s are summari zed in Table 2.
Resul t s show that as additional psychol ogi cal st at es are added t o the
regressi on equat i ons, the amount of out come measure vari ance cont rol l ed
does i ndeed i ncrease, consi st ent wi t h the model. It should be not ed, how-
ever, that the i ncrease in R 2 is substantially greater bet ween one and t wo
predi ct ors than it is bet ween t wo and three. Since measur es of t he t hree
psychol ogi cal st at es are t hemsel ves moderat el y i nt ercorrel at ed (median =
.33), and since some i ncrement in prediction is to be expect ed on purel y
statistical grounds when predi ct ors are added to a regressi on, the conclu-
T
A
B
L
E

l
M
E
D
I
A
N

C
O
R
R
E
L
A
T
I
O
N
S

O
F

J
O
B

D
I
M
E
N
S
I
O
N
S

A
N
D

P
S
Y
C
H
O
L
O
G
I
C
A
L

S
T
A
T
E
S

W
I
T
H

T
H
E

W
O
R
K

O
U
T
C
O
M
E
S
a

O
u
t
c
o
m
e

m
e
a
s
u
r
e
s

I
n
t
e
r
n
a
l

G
e
n
e
r
a
l

G
r
o
w
t
h

R
a
t
e
d

w
o
r
k

m
o
t
i
v
a
t
i
o
n

s
a
t
i
s
f
a
c
t
i
o
n

s
a
t
i
s
f
a
c
t
i
o
n

A
b
s
e
n
t
e
e
i
s
m

e
f
f
e
c
t
i
v
e
n
e
s
s

~
q

>

P
s
y
c
h
o
l
o
g
i
c
a
l

s
t
a
t
e
s

H

E
x
p
e
r
i
e
n
c
e
d

m
e
a
n
i
n
g
f
u
l
n
e
s
s

.
6
4
*
*

.
6
4
*
*

.
6
4
*
*

.
0
3

.
1
3
"

~

E
x
p
e
r
i
e
n
c
e
d

r
e
s
p
o
n
s
i
b
i
l
i
t
y

6
5
*
*

.
4
1
"
*

.
5
1
"
*

-
.
1
6

.
1
6
'
*

O

K
n
o
w
l
e
d
g
e

o
f

r
e
s
u
l
t
s

.
2
3
*
*

.
3
3
*
*

.
3
3
*
*

-
.

1
1

.
1
0

J
o
b

c
h
a
r
a
c
t
e
r
s
t
i
c
s

S
k
i
l
l

v
a
r
i
e
t
y

.
3
4
*
*

.
3
2
*
*

.
4
8
*
*

-
.

1
5
"
*

.
0
7

T
a
s
k

i
d
e
n
t
i
t
y

.
2
5
*
*

.
2
2
*
*

.
2
9
*
*

-
.

1
8

.
1
5
"
*

T
a
s
k

s
i
g
n
i
f
i
c
a
n
c
e

.
3
1
"
*

.
2
1
"
*

.
3
5
*
*

.
1
6

.
1
2
"
*

Z

A
u
t
o
n
o
m
y

.
3
1
"
*

.
3
8
*
*

.
5
1
"
*

-
.
2
4
*
*

.
1
9
"
*

O

F
e
e
d
b
a
c
k

.
3
5
*
*

.
3
8
*
*

.
4
5
*
*

-
.

1
2

.
2
1
"
*

M
o
t
i
v
a
t
i
n
g

P
o
t
e
n
t
i
a
l

S
c
o
r
e

.
4
8
*
*

.
4
3
*
*

.
5
8
*
*

-
.
2
5
"

.
2
4
*
*

C
o
r
r
e
l
a
t
i
o
n
s

w
e
r
e

c
o
m
p
u
t
e
d

s
e
p
a
r
a
t
e
l
y

f
o
r

e
a
c
h

o
f

t
h
e

s
e
v
e
n

o
r
g
a
n
i
z
a
t
i
o
n
s

w
h
e
r
e

d
a
t
a

w
e
r
e

c
o
l
l
e
c
t
e
d
,

a
n
d

m
e
d
i
a
n
s

a
r
e

r
e
p
o
r
t
e
d

h
e
r
e

S
e
e

t
e
x
t

f
o
r

e
x
p
l
a
n
a
t
i
o
n

o
f

h
o
w

l
e
v
e
l
s

o
f

s
t
a
t
i
s
t
i
c
a
l

s
i
g
n
i
f
i
c
a
n
c
e

w
e
r
e

c
o
m
p
u
t
e
d

T
o
t
a
l

n

=

6
5
8
.

*
p

<

.
0
5
.

*
*

p

<

.
0
1
.

TABLE 2
AVERAGE VARIANCE CONTROLLED IN REGRESSIONS PREDICTING OUTCOME MEASURES
FOM ONE, TWO AND THREE PSYCHOLOGICAL STATES
Mean R z for outcome measures
Number of predictors Internal General Growth
used in regressions motivation satisfaction satisfaction
One (EM; ER; KR) ~ ,29 .23 .26
Two (EM + ER; EM + KR; ER + KR) .45 .39 .43
Three (EM + ER + KR) .51 .46 ,50
n = 658. EM = experienced meaningfulness; ER = experienced responsibility; KR =
knowledge of results.
sion t hat predi ct i on is maxi mi zed when all t hree psychol ogi cal st at es are
present must be i nt erpret ed with consi derabl e caution.
Are j ob dimension-outcome variable relationships dependent on the
psychological states? Two compl ement ary met hods wer e used to t est this
quest i on. First, relationships bet ween each j ob di mensi on and t he several
out come measures wer e exami ned bef or e and after the model-specified
mediating psychol ogi cal state was statistically cont rol l ed (by partial corre-
lation). Thus, the effect of exper i enced meaningfulness was cont rol l ed for
relationships of skill vari et y, t ask identity, and t ask significance wi t h the
out come measures; experi enced responsibility was cont rol l ed for the rela-
tionships bet ween aut onomy and the out come measures; and knowl edge
of results was cont rol l ed for t he relationships bet ween f eedback and the
out come measures. If the model is correct , the partial correlations should
approach zero and be substantially l ower in magnitude than the di rect or
zero-order correlations bet ween the j ob di mensi ons and the out come
measures.
Resul t s are shown in Tabl e 3. In general, subst anual support is found
for t he proposi t i on that the psychol ogi cal st at es medi at e bet ween the j ob
di mensi ons and the out come measures. For each relationship bet ween a
j ob di mensi on and an out come measure, statistically controlling t he cor-
respondi ng psychol ogi cal state substantially l owers the magnitude of the
association. In addition, most of t he partial correl at i ons are quite l ow, and
many approach zero as predi ct ed.
Resul t s are somewhat less strong for f eedback and for aut onomy than
for the ot her j ob di mensi ons. Although relationships bet ween t hese vari-
ables and the out come measur es do decr ease moderat el y when the cor-
respondi ng psychol ogi cal st at es are cont rol l ed for, partial correlations
involving f eedback do not approach zero for any of t he dependent meas-
TABLE 3
RELATIONSHIPS BETWEEN JOB DIMENSIONS AND THE OUTCOME MEASURES CONTROLLING
FOR THE EFFECTS OF THE PSYCHOLOGICAL STATES
Zero-order Partial
Job dimension correlation correlation a Difference
Internal motivation
Skill variety .42 .15 .27
Task identity .22 .08 .14
Task significance .32 .07 .25
Autonomy .33 .08 .25
Feedback .36 .28 .08
General satisfaction
Task significance .24 - . 06 ,30
Feedback ,37 .23 .14
Growth satisfaction
Task significance .33 .06 .27
Feedback .44 .31 .13
For each job dimension, the partial correlation reported controls only for the specific
psychological state specified by the model to mediate the effects of that dimension. Thus, for
relationships involving skill variety, task identity, and task significance, experienced
meaningfulness was controlled; for relationships involving autonomy, experienced responsi-
bility was controlled; and for relationships involving feedback, knowledge of results was
controlled. (n = 658.)
ur e s , a n d pa r t i a l s i n v o l v i n g a u t o n o my a p p r o a c h z e r o o n l y f o r t he me a s -
ur e o f i nt e r na l mo t i v a t i o n .
An a ddi t i ona l a nd c o mp l e me n t a r y a n a l y s i s wa s c o n d u c t e d us i ng mul t i -
pl e r e g r e s s i o n . F o r e a c h o f t h e o u t c o me me a s u r e s , t he t h r e e p s y c h o l o g i c a l
s t a t e s we r e i n t r o d u c e d i nt o a mul t i pl e r e g r e s s i o n e q u a t i o n t o s e r v e as
p r i ma r y p r e d i c t o r s . Ne x t , t he fi ve j o b d i me n s i o n s we r e a d d e d t o t he r e-
g r e s s i o n as s e c o n d a r y p r e d i c t o r s . I f t h e p s y c h o l o g i c a l s t a t e s d o me d i a t e
t he j o b d i me n s i o n - o u t c o me me a s u r e r e l a t i ons hi ps as p r e d i c t e d , (a) t he
p s y c h o l o g i c a l s t a t e s a l one s h o u l d a c c o u n t f o r a s i z a bl e p o r t i o n o f t he
d e p e n d e n t va r i a bl e v a r i a n c e , a n d (b) i n t r o d u c t i o n o f t he fi ve j o b d i me n -
s i ons i nt o t he e q u a t i o n (as a ddi t i ona l p r e d i c t o r s ) s h o u l d n o t s ubs t a nt i a l l y
i n c r e a s e t he a mo u n t o f d e p e n d e n t va r i a bl e v a r i a n c e c o n t r o l l e d .
Re s u l t s a r e s h o wn i n Ta b l e 4. As p r e d i c t e d , t he p s y c h o l o g i c a l s t a t e s
a c c o u n t f o r s ubs t a nt i a l v a r i a n c e f o r e a c h o f t he d e p e n d e n t me a s u r e s .
Moreover, t he introduction of t he five j ob dimensions into t he regression
equations resul t ed in a near-zero i ncrease in t he vari ance controlled for
two of t he dependent measures and only a small i ncrease for t he third.
Exami nat i on of t he regressi on coefficients for t he individual variables in
t he equat i ons r eveal s a f ew anomal i es. I deal l y, t he s t andar di zed
coefficients for the psychological states woul d all be moder at e to high,
and woul d all exceed t he coefficients for t he five j ob dimensions. It was
found, however, t hat experi enced responsibility adds little to prediction
for two of t he out come measures (general and growt h satisfaction). For
bot h of t hese out come measures aut onomy (the j ob di mensi on theoreti-
cally medi at ed by experi enced responsibility) has a relatively larger re-
gression coefficient t han does experi enced responsibility. In addition, t he
coefficients for knowl edge of results are relatively small (and, for one of
t he out come measures, is trivially negative).
In sum, t he results in Tables 3 and 4 provide generally strong support
for t he predi ct i ons of t he j ob char act er i st i cs model , al t hough some
difficulties having to do wi t h cert ai n specific j ob di mensi on- psychol ogi -
cal state relationships wer e identified. Additional dat a rel evant to t hese
concerns are report ed in t he following section.
Do specific job dimensions relate to the psychological states as spec-
ified by the model? To t est this question, regressi ons wer e comput ed for
each of the psychological states, in whi ch t he predi ct ors were the j ob
dimensions specified in t he model as directly causal of that psychological
state. Thus, experi enced meaningfulness was predi ct ed from skill variety,
t ask identity, and t ask significance; experi enced responsibility was pre-
di ct ed from aut onomy; and knowl edge of results was predi ct ed from feed-
back. Next , t he remaining j ob dimensions (that is, t hose not expect ed to
di rect l y influence t he psychological state) wer e i nt roduced into each
regression equat i on as additional predictors, ff t he model is correct , t he
t heory-speci fi ed j ob dimensions should account for substantial variance
in t he psychological states, and t he i nt roduct i on of t he remaining j ob
dimensions should not substantially i ncrease t he amount of vari ance con-
trolled.
Results are present ed in Table 5, and show that a moderat e amount of
vari ance in t he psychological states is controlled by t he model-specified
j ob dimensions. For t he equations predicting experi enced meaningfulness
and knowl edge of results, t he addition of j ob dimensions not predi ct ed by
t he t heor y to affect t hese psychological states resul t ed in ver y small in-
creases in t he level of predi ct i on attained, consi st ent with t he model. The
standardized regression weights for t he equations predicting t hese t wo
variables also are as woul d be expect ed, with t he except i on of a very low
weight for t ask identity in predicting experi enced meaningfulness.
For t he equat i on predicting experi enced responsibility, results are less
T
A
B
L
E

4

M
U
L
T
I
P
L
E

R
E
G
R
E
S
S
I
O
N
S

P
R
E
D
I
C
T
I
N
G

T
H
E

O
U
T
C
O
M
E

M
E
A
S
U
R
E
S
F
R
O
M

A
L
L

P
R
I
O
R

V
A
R
I
A
B
L
E
S

C
O
M
P
A
R
E
D
T
O

P
R
E
D
I
C
T
I
O
N
S

F
R
O
M
T
H
E

P
S
Y
C
H
O
L
O
G
I
C
A
L
S
T
A
T
E
S
O
N
L
Y

a

S
u
m
m
a
r
y

s
t
a
t
i
s
t
i
c
s

R

z

f
o
r

t
h
e

t
h
r
e
e
-
v
a
r
i
a
b
l
e

R

2

f
o
r

t
h
e

f
u
l
l

I
n
c
r
e
a
s
e

i
n

R

~

b
y

a
d
d
i
n
g

M
u
l
t
i
p
l
e

c
o
r
r
e
l
a
t
i
o
n

(
R
)

f
o
r

t
h
e

e
q
u
a
t
i
o
n

(
P
s
y
c
h
o
l
o
g
i
c
a
l

e
i
g
h
t
-
v
a
r
i
a
b
l
e

t
h
e

f
i
v
e

j
o
b

d
i
m
e
n
s
i
o
n
s

f
u
l
l

e
i
g
h
t
-
v
a
r
i
a
b
l
e

e
q
u
a
t
i
o
n

s
t
a
t
e
s

o
n
l
y
)

e
q
u
a
t
i
o
n

t
o

t
h
e

r
e
g
r
e
s
s
i
o
n

I
n
t
e
r
n
a
l

m
o
t
i
v
a
t
i
o
n

.
7
2

.
5
1

.
5
2

,
0
1

G
e
n
e
r
a
l

s
a
t
i
s
f
a
c
t
i
o
n

,
6
9

.
4
6

,
4
8

.
0
2

G
r
o
w
t
h

s
a
t
i
s
f
a
c
t
i
o
n

.
7
7

,
5
0

.
5
9

.
0
9

S
t
a
n
d
a
r
d
i
z
e
d

r
e
g
r
e
s
s
i
o
n

w
e
i
g
h
t
s

(
f
o
r

t
h
e

f
u
l
l

e
q
u
a
t
i
o
n
)

E
x
p
e
r
i
e
n
c
e
d

E
x
p
e
r
i
e
n
c
e
d

K
n
o
w
l
e
d
g
e

S
k
i
l
l

T
a
s
k

T
a
s
k

m
e
a
n
i
n
g
f
u
l
n
e
s
s

r
e
s
p
o
n
s
i
b
i
l
i
t
y

o
f

r
e
s
u
l
t
s

v
a
r
i
e
t
y

i
d
e
n
t
i
t
y

s
i
g
n
i
f
i
c
a
n
c
e

A
u
t
o
n
o
m
y

F
e
e
d
b
a
c
k

I
n
t
e
r
n
a
l

m
o
t
i
v
a
t
i
o
n

.
3
1

.
4
3

-
.
0
3

.
0
9

-
.
0
1

.
0
2

-
.
0
5

,
0
8

G
e
n
e
r
a
l

s
a
t
i
s
f
a
c
t
i
o
n

.
5
2

.
0
5

.
1
2

.
0
7

-

.
0
0

-

.
0
7

.
1
0

.
0
3

G
r
o
w
t
h

s
a
t
i
s
f
a
c
t
i
o
n

.
3
8

,
0
7

.
0
9

.
1
3

.
0
3

.
0
2

.
2
4

.
0
7

a

n

=

6
5
8
.

0

Z

O

.1
8

O

t
-
a

T
A
B
L
E

5

M
U
L
T
I
P
L
E

R
E
G
R
E
S
S
I
O
N
S
P
R
E
D
I
C
T
I
N
G
T
H
E

P
S
Y
C
H
O
L
O
G
I
C
A
L
S
T
A
T
E
S
F
R
O
M

A
L
L

J
O
B

D
I
M
E
N
S
I
O
N
S
C
O
M
P
A
R
E
D
T
O

P
R
E
D
I
C
T
I
O
N
S
F
R
O
M
T
H
E

M
O
D
E
L
-
S
P
E
c
I
F
I
E
D
J
O
B

D
I
M
E
N
S
I
O
N
S
O
N
L
Y

o
o

S
u
m
m
a
r
y

s
t
a
t
i
s
t
i
c
s

M
u
l
t
i
p
l
e

c
o
r
r
e
l
a
t
i
o
n

(
R
)

f
o
r

t
h
e

f
u
l
l

e
q
u
a
t
i
o
n

(
A
l
l

f
i
v
e

j
o
b

d
i
m
e
n
s
i
o
n
s
)

R

z

f
o
r

m
o
d
e
l
-
s
p
e
c
i
f
i
e
d

j
o
b

d
i
m
e
n
s
i
o
n
s

o
n
l
y
~

R

2

f
o
r

t
h
e

f
u
l
l

e
q
u
a
t
i
o
n

(
A
l
l

f
i
v
e

j
o
b

d
i
m
e
n
s
i
o
n
s
)

I
n
c
r
e
a
s
e

i
n

R

~

b
y

a
d
d
i
n
g

t
o

t
h
e

r
e
g
r
e
s
s
i
o
n

t
h
o
s
e

j
o
b

d
i
m
e
n
s
i
o
n
s

n
o
t

s
p
e
c
i
f
i
e
d

b
y

t
h
e

m
o
d
e
l

E
x
p
e
r
i
e
n
c
e
d

m
e
a
n
i
n
g
f
u
l
n
e
s
s

E
x
p
e
r
i
e
n
c
e
d

r
e
s
p
o
n
s
i
b
i
l
i
t
y

K
n
o
w
l
e
d
g
e

o
f

r
e
s
u
l
t
s

.
6
6

.
3
8

.
4
3

.
0
5

.
5
7

.
1
7

.
3
3

.
1
6

Z

.
5
6

.
2
9

.
3
1

.
0
2

S
t
a
n
d
a
r
d
i
z
e
d

r
e
g
r
e
s
s
i
o
n

w
e
i
g
h
t
s

S
k
i
l
l

T
a
s
k

T
a
s
k

v
a
r
i
e
t
y

i
d
e
n
t
i
t
y

s
i
g
n
i
f
i
c
a
n
c
e

A
u
t
o
n
o
m
y

F
e
e
d
b
a
c
k

E
x
p
e
r
i
e
n
c
e
d

m
e
a
n
i
n
g
f
u
l
n
e
s
s

E
x
p
e
r
i
e
n
c
e
d

r
e
s
p
o
n
s
i
b
i
l
i
t
y

K
n
o
w
l
e
d
g
e

o
f

r
e
s
u
l
t
s

[
.
3
0
]

[
.
0
5
]

[
.
2
7
]

.
1
7

.
1
7

.
2
1

.
1
7

.
1
9

[
.
1
4
]

.
1
6

-
.
1
3

.
0
4

.
0
7

.
1
1

[
.
5
1
]

a

T
h
e

m
o
d
e
l
-
s
p
e
c
i
f
i
e
d

j
o
b

d
i
m
e
n
s
i
o
n
s

u
s
e
d

i
n

c
o
m
p
u
t
i
n
g

t
h
e
s
e

r
e
g
r
e
s
s
i
o
n
s

a
r
e
:

s
k
i
l
l

v
a
r
i
e
t
y
,

t
a
s
k

i
d
e
n
t
i
t
y
,

a
n
d

t
a
s
k

s
i
g
n
i
f
i
c
a
n
c
e

t
o

p
r
e
d
i
c
t

e
x
p
e
r
i
e
n
c
e
d

m
e
a
n
i
n
g
f
u
l
n
e
s
s
;

a
u
t
o
n
o
m
y

t
o

p
r
e
d
i
c
t

e
x
p
e
r
i
e
n
c
e
d

r
e
s
p
o
n
s
i
b
i
l
i
t
y
;

a
n
d

f
e
e
d
b
a
c
k

t
o

p
r
e
d
i
c
t

k
n
o
w
l
e
d
g
e

o
f

r
e
s
u
l
t
s
.

R
e
g
r
e
s
s
i
o
n

c
o
e
f
-

f
i
c
i
e
n
t
s

f
o
r

t
h
e

m
o
d
e
l
-
s
p
e
c
i
f
i
e
d

j
o
b

d
i
m
e
n
s
i
o
n
s

a
r
e

b
r
a
c
k
e
t
t
e
d

i
n

t
h
e

l
o
w
e
r

h
a
l
f

o
f

t
h
e

t
a
b
l
e
.

(
n

=

6
5
8
.
)

supportive of t he model. The vari ance controlled in this regression in-
creased . 16 (compared to .05 and .02 for t he ot her psychological states)
when j ob dimensions expect ed not to affect experi enced responsibility
were added. Moreover, exami nat i on of the regressi on weights shows that
all five of t he j ob di mensi ons cont ri but e at a moder at e level to t he predic-
tion of experi enced responsi bi l i t y. The zero-order correlations bet ween
experi enced responsibility and each of t he j ob dimensions also were
exami ned, and are consi st ent with t he regressi on findings: t he five correla-
tions are ver y homogeneous, ranging from .34 to .37 (all statistically reli-
able at less t han t he .01 level).
In sum, t he results report ed above show that t he j ob dimensions predict
experi enced meaningfulness and knowl edge of results generally as woul d
be expect ed from t he j ob charact eri st i cs model. Exper i enced responsibil-
ity, however, t urns out to be almost equally affect ed by all of the j ob
di mensi ons- - not j ust by aut onomy, as specified by t he model.
Test of the Moderating Effect of Growth Need Strength
The j ob charact eri st i cs model specifies that individual growt h need
st rengt h (GNS) can moder at e empl oyees' react i ons to t hei r work at two
points in t he motivational sequence present ed in Fig. 1. In particular, it is
predi ct ed (a) that t he relationship bet ween t he t hree psychological states
and t he out come variables will be st ronger for individuals with high
growt h need st rengt h than for individuals with low need for growth; and
(b) that t he relationship bet ween t he core j ob charact eri st i cs and their
correspondi ng psychological states will be st ronger for high t han for low
GNS individuals. In effect, t he predictions are that high GNS individuals
will be bot h bet t er able to experi ence t he psychological effects of an
obj ect i vel y enri ched j ob, and mor e disposed to respond favorabl y to that
experi ence.
The measure of growt h need st rengt h used to test t hese predictions was
obtained from t he "j ob choi ce" section of t he JDS (Hackman & Oldham,
1975). Briefly, respondent s indicate t hei r relative preference for 12 pairs
of hypot het i cal j obs (e.g., " A j ob where you are oft en requi red to make
i mport ant deci si ons" vs " A j ob wi t h many pl easant peopl e to wor k
wi t h"). For each item, a j ob with characteristics rel evant to growt h need
satisfaction is paired with a j ob having the potential for satisfying one of a
vari et y of ot her needs. Based on their scores on this measure, the top and
bot t om quartiles of empl oyees in each organization were identified, and
appropriate correlations were comput ed separat el y for t hese two groups.
For each relationship tested, it was predi ct ed that t he correlation would
be higher for empl oyees in the top quartile of t he distribution of GNS
scores than for t hose in t he bot t om quartile.
To test t he moderat i ng effects of GNS on t he psychological s t at e- out -
come measure relationship, it was desi rabl e to use a single measure that
woul d summari ze the degree to whi ch all t hree psychol ogi cal st at es simul-
t aneousl y are present. The pr oduct of the t hree psychol ogi cal st at es has
this propert y, and t herefore was correl at ed with each out come measure,
separat el y for subj ect s high and l ow in measur ed GNS. The t op group of
correlations in Tabl e 6 show t he results. Except for the measure of absen-
teeism, di fferences in the magnitude of the correlations for high vs l ow
GNS empl oyees are all in the predi ct ed direction and statistically sig-
nificant.
The r el at i ons hi ps b e t we e n t he cor e j o b char act er i s t i cs and t he
TABLE 6
RELATIONSHIPS AMONG JOB DIMENSIONS, PSYCHOLOGICAL STATES, AND OUTCOME
MEASURES FOR EMPLOYEES HIGH AND LOW IN GROWTH NEED STRENGTH (GNS)a
Median correlations
Low GNS High GNS
z (for difference
between rs)
Product of the three psychological states with:
Internal motivation .48 .66 1.75*
General satisfaction .36 .69 3.66**
Growth satisfaction .42 .69 2.68**
Absenteeism - . 16 - . 13 -0. 21
Rated work effectiveness .12 .44 2.06*
Job dimensions with corresponding psychological states
MPS with product of the psychological states .59 .70 2.02*
Skill variety with experienced
meaningfulness .23 .57 3.37**
Task identity with experienced
meaningfulness .17 .30 1.08
Task significance with experienced
meaningfulness .15 .52 2.18"
Autonomy with experienced responsibility .11 .59 2.99**
Feedback with knowledge of results .42 .63 2.54**
Motivating potential score with:
Internal motivation .27 .52 1.64*
General satisfaction .32 .49 0.93
Growth satisfaction .55 .65 0.52
Absenteeism - .23 - .25 0.00
Rated work effectiveness .20 .44 0.53
Correlations were computed separately for each of the seven organizations, and medians
are reported here. Statistical significance of the differences between correlations for high
and low GNS subjects was determined by combining the p values obtained in the separate
analyses (Mosteller & Bush, 1954, p. 329). Total n = 356 (170 and 186, respectively, in the
high and low GNS groups; ns are unequal because of tied scores).
*p < .05.
** p < .01.
MOTI VATI ON THROUGH DESI GN OF WORK 271
psychol ogi cal states for high vs l ow GNS empl oyees are shown in the
middle group of correlations in Table 6. Incl uded is a summary relation-
ship bet ween the overall MPS of the j ob and the pr oduct of the psychol og-
ical states. All di fferences bet ween correl at i ons are in the predi ct ed direc-
tion and (except for t ask identity) are statistically significant.
The bot t om group of correl at i ons in Tabl e 6 shows results for correla-
tions comput ed directly bet ween the overall motivating potential of the
j ob and the out come measures, in effect, bridging the mediating function
of the psychol ogi cal states. Again, all di fferences bet ween correlations for
high vs l ow GNS empl oyees are in the predi ct ed direction, but the differ-
ences are less substantial t han the ot hers report ed in the table, and statis-
tical significance is achi eved onl y for the measur e of internal motivation.
DISCUSSION
Empirical Validity of the Job Characteristics Model
The results report ed above provi de generally strong support for the
validity of the j ob charact eri st i cs model. A number of specific probl ems
and uncert ai nt i es were identified, however , and are expl ored bel ow.
The basic relationships bet ween the j ob dimensions and the out come
measures (Table 1) were as predi ct ed and generally of substantial mag-
nitude, although correlations involving absent eei sm and wor k perform-
ance were l ower than t hose for the ot her out come measures. Similarly,
substantial support was found for t he proposi t i on t hat individual growt h
need strength moder at es ot her model-specified relationships, and that the
moderating effect occurs bot h at the link bet ween the j ob dimensions and
the psychol ogi cal st at es, and at the link bet ween the psychol ogi cal states
and the out come variables (Table 6). This moderating effect was not,
however , obt ai ned for the measure of absent eei sm.
Bot h subst ant i ve and met hodol ogi cal explanations are possible for the
relative weakness of the results involving absent eei sm (and, to some ex-
tent, wor k performance). At the subst ant i ve level, it may be that t hese
behavi oral out comes are in fact more causal l y remot e from j ob charac-
teristics than are empl oyees' affective react i ons to their work, and there-
fore are less powerful l y affect ed by the j ob dimensions. Or t he explana-
tion may lie in the fact that the mot i vat i on and satisfaction items were in
the same questionnaire as the items tapping the j ob dimensions and the
psychol ogi cal states. For this reason, relationships involving the affect i ve
measures may have been inflated because of common met hod variance,
causing the resul t s for absent eei sm and performance t o appear weaker by
compari son. Mor eover , relationships involving performance and absen-
t eei sm may have been at t enuat ed because of the difficulty in obtaining
measur es of t hese criteria that were at the same time psychomet ri cal l y
adequat e and comparabl e across the di versi t y of j obs and organizations
studied. Finally, the results for absent eei sm may have been compr omi sed
to some ext ent by a rat her prosai c dat a collection difficulty. Because of
idiosyncratic pr ocedur es for collecting and recordi ng absent eei sm dat a in
some organizations, it was necessar y to code all absences in terms of the
number of days individuals wer e absent in a year (rather t han t he number
of occasions t hey wer e absent , as originally intended). As a result, when
an individual was away from wor k for a large number of cont i guous days
(perhaps because of a single seri ous illness or ot her personal emergency),
t hat per son woul d recei ve a ver y high absent eei sm score, when in fact the
per son may ot herwi se have had perfect at t endance. This dat a collection
probl em may have compr omi sed the overall validity of t he absent eei sm
measure used in the research. Unfort unat el y, t he present dat a do not
permit t est of t he degree t o whi ch the vari ous explanations offered above
are responsi bl e for the apparent at t enuat i on of the relationships involving
absent eei sm and wor k performance effect i veness.
Resul t s present ed in Tabl es 2 t hrough 5 provi de general (and somet i mes
quite strong) support for the proposi t i on that t he effect s of the core j ob
di mens i ons on t he o u t c o me var i abl es ar e me di a t e d by t he t hr ee
psychol ogi cal states. The onl y not ewor t hy anomalies identified are t hat
(a) results involving the f eedback dimension are in some cases less strong
t han t hose obt ai ned for the ot her j ob di mensi ons; and (b) t he aut onomy-
experi enced responsibility linkage does not oper at e as specified by the
model in predicting t he out come variables.
The probl em wi t h f eedback is not a seri ous one, and may have resul t ed
because t he pr esent st udy dealt only wi t h f eedback that deri ved from the
j ob itself. Obvi ousl y, f eedback is r ecei ved by empl oyees from many addi-
tional sources: supervi sors, peers, and so on. Mor eover , there is reason to
bel i eve that f eedback from vari ous sour ces may interact wi t h one anot her
in affecting individuals' knowl edge of the results of their wor k and their
affect i ve react i ons to t he j ob as a whol e (Greller, Not e 2). Therefore, it may
be that t he present resul t s showing how f eedback affects the out come
measures via the psychological states are, in themselves, accurate.--but that
t he resul t s are not as strong as t hey might be because f eedback from ot her
(nonjob) sour ces was not account ed for in the anal yses.
The difficulty wi t h the aut onomy- exper i enced responsi bi l i t y linkage is
more serious, because it raises quest i ons about the validity of part of the
model itself. Resul t s showed t wo findings t hat wer e cont rary t o expect a-
tion: (a) experi enced responsi bi l i t y is det ermi ned not onl y by aut onomy but
by ot her j ob di mensi ons as well (Table 5), and (b) aut onomy has direct
effect s on certain of the out come vari abl es t hat equal or exceed its pre-
di ct ed indirect i mpact via exper i enced responsibility (Tables 3 and 4).
These results do not cast doubt on t he desirability of high aut onomy and
high experi enced responsi bi l i t y for achieving beneficial wor k out comes;
t he i mpact of bot h vari abl es on t he out come measur es is, as predi ct ed,
positive. But the findings do raise quest i ons about the causal dynami cs by
whi ch such effect s are realized.
The explanation for t hese anomalies may deri ve part l y from the rela-
tionships among the j ob dimensions t hemsel ves. The five dimensions are
not empirically i ndependent (Hackman and Oldham, 1975, report a me-
dian intercorrelation of .26), nor woul d t hey be expect ed to be: Jobs that
are " good" oft en are good in several ways, and j obs that are " b a d " oft en
are generally bad. It also is the case, however , t hat aut onomy is the least
i ndependent of the five j ob dimensions (the medi an correl at i on of au-
t onomy wi t h the ot her di mensi ons is .36). Thus, it may be that aut onomy
serves, at least in part, to summari ze the overall compl exi t y of a j ob, and
that it t herefore is bot h more multiply det ermi ned and has a great er diver-
sity of effect s than do the ot her j ob dimensions. If this is the case, of
course, the funct i ons of bot h aut onomy and experi enced responsibility in
the model-specified causal sequence woul d be empirically muddied. Addi-
tional research will be requi red to obt ai n i ncreased specificity and clarity
regarding the funct i ons of aut onomy and experi enced responsibility in the
j ob charact eri st i cs model.
MPS as a Summary Measure of the Job Characteristics
The Mot i vat i ng Potential Score (MPS) has been used t hroughout this
paper as a devi ce for summarizing the overall degree to whi ch a j ob is
obj ect i vel y designed in a way that maximizes the possibility for internal
mot i vat i on on the part of the peopl e who perform it. The MPS formulation
deri ves directly from the proposi t i ons of the j ob characteristics model,
and t herefore shoul d be valid t o the ext ent that t he model itself has valid-
ity. Yet it is i mport ant to compar e the empirical performance of MPS with
that of simpler alternative model s because t here is increasing evi dence
that in a wi de vari et y of predi ct i on situations simple, unwei ght ed linear
model s out perform more compl ex and subtle formulations (cf. Dawes &
Corrigan, 1974). Mor eover , special concern about MPS may be war-
rant ed, because the MPS formul a includes t wo multiplicative terms.
Gi ven that multiplicative operat i ons can compound the effect s of measure
unreliability (and are rarel y warrant ed in any case by the scale propert i es
of the data), t here is cause for concern about how MPS predi ct i ons com-
pare to t hose based on nonmultiplicative models.
Fi ve different model s for combining the j ob di mensi ons were devel oped
and correl at ed with the t hree quest i onnai re-based dependent measures.
The five model s and the correl at i ons obt ai ned are shown in Table 7. The
results do not meaningfully differentiate among the models. While the full
multiplicative model pr oves to be slightly the worst , and the regressi on
model s are slightly the best , the obt ai ned di fferences are so small as to be
of negligible practical significance. Thus, while the model-specified MPS
formulation is not disconfirmed by the data, neither has it been shown to
represent a more adequat e means of combining the j ob dimensions than
ot her, simpler alternatives.
The Nature and Effects of Growth Need Strength
Some researchers (e.g., Hackman, Oldham, Janson, & Purdy, 1975;
Hulin & Bl ood, 1968) have suggested that individuals who are l ow in
gr owt h need st rengt h (or who are al i enat ed from mi ddl e-cl ass wor k
norms) may react negatively to compl ex or enri ched j obs because t hey
will be psychol ogi cal l y " s t r et ched" t oo far by such j obs, or because t hey
will not value the kinds of out comes that such j obs provide.
The present findings provi de no evi dence to support such cont ent i ons.
While individuals wi t h strong growt h needs do react more posi t i vel y t o
compl ex j obs t han do individuals wi t h weak needs for growth, t he signs of
the relationships bet ween the j ob characteristics and the out come meas-
ures are posi t i ve even for peopl e in the bot t om quartile of t he growt h
need measure. This is of special significance in t he present st udy, because
the sampl e included several groups of empl oyees who scored especially
l ow on the measure of growt h need strength.
Such individuals may not be pri med and r eady t o respond enthusiasti-
cally to a j ob that is more compl ex and challenging t han the one t hey now
hold. For this reason, t hose responsi bl e for the implementation of j ob
enri chment programs might be wel l -advi sed to pr oceed sl owl y and care-
TABLE 7
COMPARISON OF SEVERAL MODELS FOR COMBINING TIlE JOB DIMENSIONS a
Out come measures
Internal General Growt h
Alternative models mot i vat i on satisfaction satisfaction
MPS [ ( S V+ TI3 + T S ' ) x Ax F] .46 .49 .63
Full multiplicative
[SV x TI x TS x A F] .44 .45 .58
Simple additive
[SV + TI + TS + A + F] .51 .52 .67
Multiple regression .52 .53 .69
Cross-validated regression .52 .53 .68
a n (except for cross-validated regression) = 658; n for cross-validated regression = 329.
fully when the target empl oyees have onl y weak needs for personal
growth. And the magnitude of the gains realized in such ci rcumst ances
may well turn out t o be less than woul d be the case for empl oyees high in
growt h need strength. But the present findings provi de no reason to ex-
pect that the ultimate i mpact of worki ng on enriched j obs will be more
negative t han positive for any group of empl oyees, regardl ess of their
level of growt h need strength (see also Stone, 1976).
The present results confirm that the moderat ors of individuals' reac-
tions to their wor k can be usefully concept ual i zed and measured directly
in terms of human needs. Quest i ons remain, however , regarding the
relationships bet ween such measures and the demographi c and subcul-
tural variables that also have been pr oposed as moderat ors.
To exami ne this issue, a summar y measure of growt h need strength was
correl at ed with a number of demographi c and background characteristics
of empl oyees in the present sample. Resul t s are shown in Tabl e 8, and
suggest that the "t ypi cal " high growt h need empl oyee is a young and
wel l -educat ed male who wor ks or lives in a suburban or rural setting. It is
not ewor t hy that the individual' s present place of wor k and resi dence
relate most substantially to measur ed need for growth, whereas the loca-
tion of socialization is rat her weakl y associ at ed (cf. Hulin & Bl ood, 1968;
Turner & Lawr ence, 1965; Wanous, 1974). Evi dent l y current experi-
ences are more responsi bl e for determining an individual' s desire for
growt h satisfaction than are items of personal history, and t herefore are
more likely to moder at e the relationships bet ween j ob charact eri st i cs and
out come variables.
If this concl usi on is accept ed, t hen r esear ch examining the effect s
of j ob and organizational st ruct ures on empl oyee gr owt h needs may
prove informative. It may be that individuals' needs change or adj ust
to meet the demands of t he situation in whi ch t hey find t hemsel ves.
Thus, the needs of an individual may actually become more "gr owt h
or i ent ed" when he is confront ed with a compl ex j ob whi ch seems to
demand that the individual devel op hi msel f and exerci se i ndependent
thought and action in his work.
Uses and Distinguishing Features of the Job Characteristics Model
The j ob charact eri st i cs model was designed so that each of the t hree
focal classes of variables (i.e., obj ect i ve j ob charact eri st i cs, mediating
psychol ogi cal states, and individual growt h need strength) can be directly
measur ed in actual wor k situations using the J ob Diagnostic Sur vey
( Hackman & Oldham, 1975). Therefore, t he model can be used as a con-
cept ual basis for the diagnosis of j obs being consi dered for redesign (e.g.,
to det ermi ne the existing pot ent i al of a j ob for engenderi ng internal
TABLE 8
RELATIONSHIPS BETWEEN GROWTH NEED STRENGTH AND
DEMOGRAPHIC CHARACTERISTICS a
Rural Rural Rural
Growth need place current childhood
strength Sex Age Education of work residence residence
Growth need strength
Sex -. 26 - -
Age -. 15 .02 - -
Education .46 -. 33 -. 05
Rural place of work .16 -. 39 -, 05
Rural current residence .14 -. 22 .01
Rural childhood residence .02 -. 16 -. 03
m
.19
.09 .35 - -
-. 02 .33 .45
a See Hackman & Oldham (Note 3) for details regarding the measurement of each
variable. For psychometric reasons, tests of significance are not appropriate for these cor-
relations. If such tests were appropriate, all correlations ~ .09 would be significant at the
.05 level, n = 658.
wo r k mot i vat i on, t o i dent i f y t hos e speci f i c j o b char act er i s t i cs t hat ar e
mo s t in ne e d of i mp r o v e me n t , and t o a s s e s s t he " r e a d i n e s s " of e mp l o y e e s
t o r e s pond pos i t i vel y t o e nr i c he d wor k) . I n addi t i on, t he mode l can s er ve
as a f r a me wo r k f or as s es s i ng and i nt er pr et i ng me a s u r e me n t s col l ect ed t o
e va l ua t e t he ef f ect s of c ha nge s t hat ha ve b e e n car r i ed out (e. g. , t o det er -
mi ne whi c h j ob di me ns i ons di d and di d not change, t o a s s e s s t he i mpa c t o f
t he c ha nge s on t he af f ect i ve and mot i va t i ona l r e s pons e s o f e mp l o y e e s ,
and t o t es t f or a ny pos s i bl e pos t c ha nge al t er at i ons in t he gr owt h need
s t r engt h of t he e mp l o y e e s wh o s e j o b s we r e r edes i gned) .
The j ob di me ns i ons speci f i ed by t he mode l ar e di r ect l y t i ed t o a set o f
act i on pr i nci pl es f or r edes i gni ng j obs ( Ha c k ma n , Ol dha m, J a n s o n &
Pur dy, 1975; Wa i t e r s & As s o c i a t e s , 1975). Th e s e pr i nci pl es s pe c i f y
wha t t ype s of c ha nge s ar e mo s t l i kel y t o l ead t o i mp r o v e me n t s in e a c h of
t he five c or e di me ns i ons , and t h e r e b y t o an ove r a l l i nc r e a s e in t he
mot i vat i ng pot ent i al of a j ob. The us ef ul nes s o f t he act i on pr i nci pl es f or
i ncr eas i ng t he MPS o f a j o b has not ye t b e e n empi r i cal l y t es t ed, h o we v e r ;
nei t her has t he val i di t y o f t he j o b char act er i s t i cs mode l i t sel f be e n as s es -
sed i n an act ual c ha nge pr oj ect . The r e f or e , f ur t he r r e s e a r c h is r equi r ed
be f or e mo r e t han t ent at i ve s t a t e me nt s c a n be ma de r egar di ng t he usef ul -
nes s of t he mode l as a pr act i cal gui de f or wo r k r edes i gn.
I t s houl d be not e d t hat t he j ob char act er i s t i cs mode l deal s onl y wi t h
a s pe c t s o f j obs t hat c a n be al t er ed t o c r e a t e positive mot i vat i onal i ncen-
t i ves f or t he j ob i nc umbe nt . I t doe s not di r ect l y a ddr e s s t he dys f unct i onal
a s pe c t s o f r epet i t i ve wo r k (as does act i vat i on t he or y) , al t hough p r e s u ma -
bl y a j ob des i gned in a c c or d wi t h t he di ct at es o f t he mode l woul d not t ur n
out t o be r out i ne or hi ghl y r epet i t i ve.
MOTI VATI ON THROUGH DESIGN OF WORK 277
In addition, the model f ocuses excl usi vel y on the relationship bet ween
i ndi vi dual s and t hei r wor k. It does not addr es s di r ect l y i nt er per -
sonal, technical, or situational moder at or s of how peopl e react t o their
wor k (as does socio-technical syst ems theory), even t hough at t ent i on to
such fact ors may be critical in successful installations of actual wor k
changes (Davis & Tayl or, 1972; Hackman, 1975). A recent st udy by Old-
ham (in press), for exampl e, has shown that inclusion of one such mod-
erat or (the quality of i nt erpersonal relationships on the j ob) significantly
i mproves prediction of empl oyees' responses to their j obs. Specifically, it
was found that peopl e who wor k on compl ex j obs experi ence great er
internal motivation when t hey are satisfied with on-t he-j ob relationships
than when t hey are dissatisfied wi t h t hese relationships.
Finally, the j ob charact eri st i cs model is designed to appl y onl y to j obs
that are carried out more-or-less i ndependent l y by individuals. It offers no
explicit gui dance for the effect i ve design of wor k for interacting t eams,
i.e., when t he wor k is best concei ved of as a group task, as is somet i mes
the case when " aut onomous wor k gr oups" are formed (Gul owsen, 1972).
The model should, nevert hel ess, be of some use in designing t asks that are
motivating to group members: Presumabl y, a " g o o d " j ob for a group
woul d have many of the same obj ect i ve charact eri st i cs as a well-designed
j ob intended for an individual.
Yet it also appears that it woul d be necessar y to go well beyond the
present limits of the j ob charact eri st i cs model in designing group t asks, for
at l east t wo reasons. First, it seems doubt ful that translating the core j ob
dimensions from the individual to the group level woul d be an entirely
straightforward process. How, for exampl e, should a group t ask be de-
signed so that all member s woul d see it as providing high aut onomy, and
t herefore experi ence substantial personal responsi bi l i t y for the out come
of the group? A second probl em deri ves from the fact that how group
tasks are designed affects not onl y the mot i vat i on of group members, but
t he pat t er ns of soci al i nt er act i on t hat emer ge among t hem as wel l
( Hackman & Morris, 1975). How can group t asks be st ruct ured so that
t hey prompt t ask-effect i ve rat her than dysfunct i onal pat t erns of interac-
tion among member s? Although such quest i ons are crucial to the effect i ve
design of t asks for t eams, t hey appear to have no simple answers, nor are
t hey quest i ons for whi ch the j ob charact eri st i cs model in its present form
provi des explicit guidance.
REFERENCES
Berl yne, D. E. Ar ousal and r ei nf or cement . Nebraska Symposium on Motivation, 1967, 15,
1- 110.
Bl ood, M. R., & Hulin, C. L. Al i enat i on, envi r onment al char act er i st i cs, and wor ker re-
sponses. Journal of Applied Psychology, 1967, 51, 284- 290.
Brief, A. P., & Aldag, R. J . Empl oyee r eact i ons t o j ob charact eri st i cs: A const r uct i ve
repl i cat i on. Journal of Applied Psychology, 1975, 60, 182-186.
Davis, L. E., & Taylor, J. C. Design of jobs. Middlesex, England: Penguin, 1972
Dawes, R. M., & Corrigan, B. Linear models in decision making. Psychological Bulletin,
1974, 81, 95-106.
Dunnette, M. D., Campbell, J. P., & Hakel, M. D. Factors contributing to job satisfaction
and dissatisfaction in six occupational groups. Organizational Behavior and Human
Performance, 1967, 2, 143-174.
Emery, F. E., & Trist, E. L. Socio-technical systems. In F. E. Emery (Ed.), Systems
thinking. London: Penguin, 1969.
Ford, R. N. Motivation through the work itself. New York: American Management Associa-
tion, 1969.
Gulowsen, J. A measure of work group autonomy. In L. E. Davis & J. C. Taylor (Eds.),
Design of jobs. Middlesex, England: Penguin, 1972.
Hackman, J. R. On the coming demise of job enrichment. In E. L. Cass and F. G. Zimmer
(Eds.), Man and work in society, New York: Van Nostrand-Reinhold, 1975.
Hackman, J. R., & Lawler, E. E. Employee reactions to job characteristics. Journal of
Applied Psychology Monograph, 1971, 55, 259-286.
Hackman, J. R., & Morris, C. G. Group tasks, group interaction process, and group
performance effectiveness: A review and proposed integration. In L. Berkowitz (Ed.),
Advances in experimental social psychology. New York: Academic Press, 1975. Vol. 8.
Hackman, J. R., & Oldham, G. R. Development of the Job Diagnostic Survey. Journal of
Applied Psychology, 1975, 60, 159-170.
Hackman, J. R., & Oldham, G., Janson, R., & Purdy, K. A new strategy for job enrichment.
California Management Review, Summer, 1975, 57-71.
Herbst, P. G. Autonomous group functioning. London: Tavistock, 1962.
Herzberg, F. Work and the nature of man. Cleveland: World, 1966.
Herzberg, F., Mausner, B., & Snyderman, B. The motivation to work. New York: Wiley,
1959.
Hinton, B. L. An empirical investigation of the Herzberg methodology and two-factor
theory. Organizational Behavior and Human Performance, 1968, 3, 286-309.
Hulin, C. L. Individual differences and job enrichment. In J. R. Maher (Ed.) New perspec-
tives in job enrichment. New York: Van Nostrand-Reinhold: 1971.
Hulin, C. L., & Blood, M. R. Job enlargement, individual differences, ana worker responses.
Psychological Bulletin, 1968, 69, 41-55.
King, N. A clarification and evaluation of the two-factor theory of job satisfaction.
Psychological Bulletin, 1970, 74, 18-31.
Lawler, E. E., & Hall, D. T. The relationship of job characteristics to job involvement,
satisfaction and intrinsic motivation. Journal of Applied Psychology, 1970, 54,
305-312.
Mosteller, F., & Bush, R. L. Selected quantitative techniques. In G. Lindzey (lEd.), Hand-
book of social psychology. Reading, Mass.: Addison-Wesley, 1954. Vol. 1.
Oldham, G. R. Job characteristics and internal motivation: The moderating effect of inter-
personal and individual variables. Human Relations. In press.
Paul, W. J. Jr., Robertson, K. B., & Herzberg, F. Job enrichment pays off. Harvard Busi-
ness Review, 1969, 47, 61-78.
Porter, L. W., Lawler, E. E., & Hackman, J. R. Behavior in organizations. New York:
McGraw-Hill, 1975.
Rice, A. K. Productivity and social organization: The Ahmedabad experiment. London:
Tavistock, 1958.
Robey, D. Task design, work values, and worker response: An experimental test. Organiza-
tional Behavior and Human Performance, 1974, 12, 264-273.
Scott, W. E. Activation theory and task design. Organizational Behavior and Human Per-
formance, 1966, ], 3-30.
Shepard, J. M. Functional specialization, alienation, and job satisfaction. Industrial and
Labor Relations Review, 1970, 23, 207-219.
Stone, E. F. The moderating effect of work-related values on the job scope-job satisfaction
relationship. Organizational Behavior and Human Performance, 1976, 15, 147-179.
Stouffer, S. A. et al. The American soldier. Vol. 1. Adjustment during army life. Princeton:
Princeton Univ. Press, 1949.
Thayer, R. E. Measurement of activation through self-report. Psychological Reports, 967,
20, 663-678.
Trist, E. L., Higgin, G. W., Murray, H., & Pollock, A. B. Organizational choice. London:
Tavistock, 1963.
Turner, A. N., & Lawrence, P. R. Industrial jobs and the worker. Boston: Harvard
Graduate School of Business Administration, 1965.
Waiters, R. W. & Associates. 3ob enrichment for results. Reading, Mass.: Addison-Wes-
ley, 1975.
Wanous, J. P. Individual differences and reactions to job characteristics. Journal of Applied
Psychology, 1974, 59, 616-622.
Whitsett, D. A., & Winslow, E. K. An analysis of studies critical of the motivator-hygiene
theory. Personnel Psychology, 1967, 20, 391-415.
REFERENCE NOTES
l. Davis, L. E., & Trist, E. L. Improving the quality of work life: Experience of the
socio-technical approach. Background paper commissioned by the U. S. Department
of Health, Education, and Welfare for the Work in America Project, June, 1972.
2. Greller, M. The consequences of feedback. Unpublished doctoral dissertation, Yale
University, 1975.
3. Hackman, J. R., & Oldham, G. R. The Job Diagnostic Survey: An instrument for the
diagnosis of jobs and the evaluation of job redesign projects. Technical Report No. 4,
Department of Administrative Sciences, Yale University, 1974. (Also available from
Journal Supplement Abstract Service of the American Psychological Association, Ms.
No. 810, 1974.)
4. Sims, H. P., & Szilagyi, A. D. Individual moderators of job characteric relationships.
Unpublished manuscript, Graduate School of Business, Indiana University, 1974.
RECZIVED: April 28, 1975

Motivation Through The Design of Work: Test of A Theory

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Motivation Through The Design of Work: Test of A Theory

Uploaded by

Copyright:

Available Formats

ORGANI ZATI ONAL BEHAVI OR AND HUMAN PERFORMANCE 1 6 , 2 5 0 - 2 7 9 (1976)

Motivation through the Design of Work: Test of a

You might also like