Professional Documents
Culture Documents
Impact of Personality and Intelligence On Job Satisfaction
Impact of Personality and Intelligence On Job Satisfaction
4, 293-299
Introduction tasks. The reasons for this are due to some of the following
Much of the job satisfaction literature pertaining to weaknesses in studies:
machine-paced (M/P) and self-paced (S/P) work has been 1. When job satisfaction was assessed in actual industrial
summarised by Davis (1982), Salvendy and Smith (1981); work situations, one of the following two conditions
Dunnette (1976); Locke (1976); Kornhauser (1965); prevailed:
Vroom (1964); Herzberg, Mausner and Snyderman (1959); a. Only workers on M/P work were studied.
Walker and Guest (1952). Unfortunately, even with this
extensive literature no firm conclusions can be drawn b. When studies compared job satisfaction on both M/P
regarding the ratios of job satisfaction on M/P and S/P and S/P work, work output and job content were
confounded with pacing and individual operators:
Different groups of people performed M/P and S/P
modes. Also the total work output (percentage of
i School of Industrial Engineering work standard) was not controlled across pacing
2 Department of Statistics modes. Furthermore, job contents in the two pacing
3 At the time of this study was at the School of Industrial conditions were entirely different.
Engineering; now at AT & T Long Lines, Cincinnati, Ohio
4 At the time of this study was at the School of Industrial
Engineering; now at Bell Laboratories, Holmdel, NJ 2. When job satisfaction on M/P and S/P work was assessed
s Professor Emeritus, Department of Psychological Sciences in laboratory situations the confounding effects discussed
above were either eliminated or markedly reduced.
The study was supported by the National Science Foundation However, the following two conditions usually prevailed:
research grant No APR7718695.
a. The durations of the laboratory experiments were of
The authors are grateful to the Naval Avionics Center, very short duration. Therefore, subjects studied were
Indianapolis, Indiana, for their kind co-operation in allowing neither sufficiently experienced nor skilled in task
us to perform the study in their facilities. Our thanks go to performance and were not exposed to the psychological
Gary Endicott for his assistance in the development of the effects of long-term task performance experienced by
Purdue Satisfaction Questionnaires. workers in field situations.
0003-6870/82/04 0293-07 $03.00 © 1982 Butterworth & Co (Publishers)Ltd AppliedErgonomics December 1982 293
b. Most subjects were university students with no prior 2. In the M/P condition each task element was completed
shop floor industrial work experience. Since this within a rigidly fixed temporal window; in S/P, element
population is markedly different from the population completion times could vary at the operator's discretion.
of unskilled industrial workers, results may not be The specific tasks were of two general varieties: one type
applicable. task consisted of loading small electronic parts and wires into
The purpose of the study described in this paper was to pre-punched printed circuit boards. Hand soldering was
overcome the limitations described above in order to required to fix some components to the boards. Careful
compare job satisfaction of M/P and S/P work. To do this, inspection of the boards was necessary to ensure an extremely
a statistically-balanced experiment was conducted in an low defective board count. Specific tasks within this category
actual industrial setting and the subjects involved were varied as different circuit boards were loaded. A step-by-
highly experienced industrial operators. step loading diagram was positioned in front of each operator
to show the proper sequence of board loading steps (ie, task
One potential limitation of this approach should be noted.
elements).
In this study the pacing mode was confounded with
simplified versus enriched jobs. However, this reflects the The other task variety consisted of attaching and soldering
reality of the situation in actual instances of S/P versus M/P precut wires to multi-terminal switch assemblies. This
industrial tasks. required extensive use of long-nose and diagonal cutting
pliers, soldering irons, and again, very critical visual
inspection. Over 100 wires were attached to each switch
Method assembly. Wiring directions were located in front of each
operator. Variants of this task type differed according to
Overview and tasks the particular switch assembly under construction.
The study was conducted at the Naval Avionics Center,
Indianapolis, Indiana. A segment of the Center's activities The work environment, the Naval Avionics Center, is
includes hand manufacture of small electronic assemblies quiet and spacious; it is generally rated very favourably by
such as printed circuit boards and switch modules. These employees. The work day began at 07.00 and terminated at
tasks require manual dexterity, visual acuity, and 16.30. A half-hour lunch break was allowed from 12.00-
concentration, but no physical exertion. Workers engaged 12.30.
in these tasks provided the subject population for this study.
Prior to this study, no M/P work systems had been utilised in
these manufacturing tasks and so all operators were Subjects
experienced only with S/P work. Twenty-eight female subjects (aged 2 8 - 6 4 years; ~ = 45-9,
sd = 10.3) volunteered for the experiment. Volunteers were
In the study, one of the work benches normally occupied obtained by a joint union and management effort in which
by four to six operators was replaced by a 7.3 m (24 ft) the entire population of employees engaged in manufacturing
carousel-type conveyor. During S/P portions of the study, was solicited. Volunteers were not given any special rewards
the conveyor was stationary and operators simply used it as for participation in the study other than a certificate of
a fixed work-bench. In M/P conditions, work assemblies appreciation from the experimenters which was added to
were placed on small pallets that remained stationary for their personnel file. All volunteers were well experienced
predetermined work periods (which ranged from 40-100 s) at the general tasks required in this study (one or more years).
and then indexed (toward the worker's right) to move the No attempt was made to select volunteers for special
workpiece to the next work station. The indexing time attributes or good health. Instead, a random selection of
was 4 s. 28 subjects was made from the volunteers.
The range of manufacturing tasks performed at the Center The participating subjects were naive on machine-paced
was carefully surveyed and a set of tasks with readily work but all subjects had, at least, one year experience on
separable task elements was selected. An example of a task self-paced work.
element was to attach and solder one wire to one switch
terminal. During M/P conditions, the conveyor indexed so
that operators performed only one task element during each Experimental design
work period. The same element was completed on all Volunteers were tested in eight groups of three and one
workpieces and then the next (differ6nt) task element was group of four. Groups were formed on the basis of
initiated. The order in which task elements were done was administrative arrangements not related to the experiment.
matched to that chosen by operators when they freely These arrangements reflected such considerations as vacation
performed their tasks in a S/P fashion. Thus, the tasks schedules and project commitments.
performed in S/P and M/P conditions were highly similar
except that: Each group experienced two consecutive weeks of M/P
work and two consecutive weeks of S/P work. The subjects
1. In M/P conditions, operators completed a given task were divided into two groups of 14 subjects. One group
element on all workpieces before beginning a new task received the S/P condition first, while the other group
element; in S/P, all task elements were completed on a experienced the reverse order. A pre-test session was
workpiece before starting on a new work item. Thus the administered on the Friday immediately preceding the
cycle times for S/P varied from 8 to 12 min, while for start of each four-week experimental period.
M/P tasks the times per cycle varied from 0.6 to 1 "6 min.
This represents a fairly typical industrial situation with Because of the manner in which volunteers became
M/P associated with simplified jobs and S/P associated available for this study, it was not possible to fully balance
with enlarged jobs. Thus pacing mode was confounded combinations of subject age with pacing order. However,
with job enlargement/simplification. balance was approximated as closely as possible.
Schedule of
administration (z)
Data analysis
Six measures of satisfaction were used. Satisfaction ratios Table 2: Personality trait and other predictors of satisfaction
are defined to be the ratio of satisfaction scores under S/P descriptive statistics
and M/P conditions.
A composite satisfaction measure was derived as follows. Standard
A principal components analysis on the seven satisfaction Label Variable* Mean deviation
scores indicated a strong overall satisfaction component. 1 Reservedvs Outgoing 10.0 2.4
The important variables in this component were JDS general 2 Less Intelligent vs More
satisfaction, JDS motivating potential, PSQ rate satisfaction Intelligent 7-5 1.4
and PSQ autonomy satisfaction, with each variable being 3 Affected by Feelings vs
of approximately equal importance. The composite score Emotionally Stable 15.1 4.3
was therefore defined to be a standardised average of these 4 Humble vs Assertive 9.1 3-2
variables: each variable was divided by its standard deviation, 5 Sober vs Happy-go-lucky 12.4 4.0
the results added, and the total was divided by the sum of 6 Experiment vs Conscientious 13.7 3.0
the inverses of the standard deviations. The resulting 7 Shy vs Venturesome 12.3 4.5
composite is thus standardised so that the expected value of 8 Tough-mindedvsTender-minded 12.2 2.3
the composite, under the null hypothesis that the ratios are 9 Trusting vs Suspicious 6.7 2.9
1, is also 1. 10 Practical vs Imaginative 10.0 3.1
To investigate the effect of S/P versus M/P conditions on 11 Forthright vs Shrewd 11 '0 2.6
satisfaction scores, paired sample t-tests were used to compare 12 Placid vs Apprehensive 10-9 3.0
the two means. In addition, the ratios were tested against a 13 Conservative vs
null hypothesis of one using a t-test. The composite Experimenting 6"4 2.5
satisfaction score was also tested using this second method. 14 Group Dependent vs Self-
sufficient 10" 1 3"9
Regression methods were used to construct prediction 15 Undisciplined Self-conflict vs
equations for the satisfaction ratios (including the Controlled 14.2 3.0
composite) using the variables in Table 2. A computer 16 Relaxed vs Tense 13-8 4.7
program which can calculate multiple correlations for all 17 Extroversion 11-8 3.5
possible subsets of predictor variables was used. Since the 18 Neu roticism 10.8 5" 1
program capability is limited to 20 variables, and since for 19 Lie Scale 2.7 1.6
this data set some numerical instabilities arose at this level, 20 Age 47.1 10.1
the 22 possible predictors were screened by running separat.e 21 Fatigue Ratio (feeling tone
analyses for the first 16 and last 16 variables. Note that these checklist) 1.0 0"3
sets have 10 variables in common. The best subset of eight 22 SRA 20.4 7.9
predictors was selected from each run and the program was
re-run using these variables. *Variables 1 to 16 are from 16PF; variables 17-19 are from
To determine how many variables to use in the ffmal the Eysenck Personality Inventory; Variable 21 is from
equation, Cpmethodology was used (see Hocking, 1976, for Fatigue checklist. References for the measures discussed in
details). In each case the size of the subset was the smallest this table are illustrated in Table 1.
Table 3: Descriptive statistics for satisfaction scores under self and machine-paced conditions
JDS General Satisfaction 3.03 "57 2.76 .69 2.54** 1.14 .26 2.85**
JDS Motivating Potential 86.91 63.96 48.02 43.68 4.11 * * 2-58 1.63 5.22"*
JDI Work Satisfaction 23.00 10"56 20.66 9.26 1.74" 1-17 0.46 2.01 *
PSQ Rate Satisfaction 4.94 1"23 3.54 0.70 6.69** 1.44 .44 5.43**
PSQ Autonomy Satisfaction 13"86 7.81 12.31 6.45 2.57** 1-17 .37 2.54**
PSQ Task Satisfaction 22.83 4-80 20"55 3.85 1.84* 1.19 .59 1.74"
Multiple correlation
The true values of the multiple correlations probably lie somewhere between the corrected values portrayed in the reduced and
the original set. These values could conceivably be further reduced in cross-validation.
CELL
BIOCHEMIST/W
FIN/) FIINC'TION
Editor: P r o f e s s o r E. M. C r o o k
(Medical College of St. Bartholomew's Hospital, UK)