Effects of individual characteristics, organizational factors and task characteristics on computer programmer productivity and job satisfaction

Effects of individual characteristics, organizational factors and task characteristics on computer programmer productivity and job satisfaction

209 Research Effects of Individual Characteristics, Organizational Factors and Task Characteristics on Computer Programmer Productivity and Job Sati...

586KB Sizes 0 Downloads 72 Views

209

Research

Effects of Individual Characteristics, Organizational Factors and Task Characteristics on Computer Programmer Productivity and Job Satisfaction 1. Introduction

Several individual characteristics of computer programmers (self- esteem. level of experience. and mathematical aptitude); four organizational factors (perceived supervisory initiated structure. perceived supervisor consideration. perceived level of performance feedback, and perceived degree of participation in organizational decisions): and five task characteristics (skull variety. task variety, task significance, autonomy, and feedhack) are related to computer programmer productivity and joh via queationsatisfaction. Measurement was accomplished natres; least squares multiple regression was used to test the hypotheses. K~~wvo~dx Computer ity.

Programmer,

I

I

Job Satisfaction,

Productiv-

Paul H. Cheney is an associate profession of management in the College of Business Administration at The University of Georgia. He obtained his Ph.D. in Management Information Systems from the University of Minnesota. His current research interest include decision support systems, implementation strategies for new information systems and measuring productivity and satisfaction among computer professionals.

Today organizations spend billions of dollars on the development and maintenance of software. Several authors [19.20,25] indicate that information systems departments now devote a greater proportion of their annual budgets to software costs than to any other category. Computer programmer salaries account for 80 to 90% of these software costs. The research discussed here might provide help in two related areas: 1. increasing the level of job satisfaction of computer programmers, and thus 2. increasing their productivity. Several characteristics relating to the individual (self-esteem, mathematical aptitude, level of experience); several organizational factors (perceived supervisory initiated structure and consideration, participation in decisions and performance feedback); and five task characteristics (skill variety, task variety, task significance, autonomy, and feedback) were related to computer programmer productivity and job satisfaction. It is apparent why productivity was chosen as a dependent variable in this research, but ~ Why job satisfaction? Job satisfaction was included because it has been inversely correlated with turnover [6,16]: and absenteeism [4,24]. Because there are significant costs incurred by an organization when these two problems exist, job satisfaction was included as a dependent variable. 2. Individual Characteristics

North-Holland Information & Management 0378.7206/84/$3,00

7 (1984) 2099214

‘i; 1984, El.sevier Science Publishers

Several individual characteristics have been linked theoretically and empirically to computer proB.V. (North-Holland)

grammer productivity and job satisfaction. They include self-esteem, level of experience. and mathematical aptitude. Based upon research by Korman [17], Lawler [18]. and Sims, et al. [22], it appears that productivity and job satisfaction are positively related to the employee’s level of selfesteem. If this is true in the case of industrial salesmen, purchasing agents, accountants. and other job classifications it may also be true for computer programmers: therefore. this variable was included as an independent variable. Specific self-esteem, as used in this research was defined by Bagozzi [2] as: particular self-attributions and self-regard with respect to actual performance on the job (e.g. perception of ability to achieve one’s potential. self-confidence, etc. , . ) A second individual characteristic variable, experience, was included in the study to test the hypothesis that the computer programmer’s job tenure is positively related to their productivity and level of job satisfaction. Previous research tends to support this hypothesis [7,9]. Level of experience, as used here, refers to the number of years the person has been employed as a computer programmer. Mathematical ability (the person’s ability to solve mathematical and logical problems) has long been thought to be positively related to computer programmer productivity by practitioners and academicians. The relationship of this variable to job satisfaction is less clear, but I felt the relationship warranted investigation and therefore it was included in the study. The three hypotheses concerning individual characteristics are: Hl: The computer programmer’s perception of specific self-esteem is related positively to his/her level of productivity and job satisfaction. H2: The computer programmer’s years of experience as a programmer is related positively to the magnitude of his/her productivity and job satisfaction. H3: The computer programmer’s mathematical aptitude is related positively to the magnitude of his/her productivity and job satisfaction.

vides consideration may be related to the employees’ level of productivity. Initiurion ofstructure is the degree to which the leader structures and defines his or her role and the role of subordinates in job-related activities, such as specifying procedures and assigning tasks [lo]. Considerution is the degree to which the leader develops a work climate of psychological support, mutual trust and respect, helpfulness and friendliness [lo]. Evans [8] found the leader’s consideration and initiation of structure to be positively related to job satisfaction and productivity in the case of utility workers Sims et al. [22] found similar results among medical center personnel. The programmer’s leader in this research was the project leader. Lawler [1X] hypothesized that jobs which provide feedback to the employees and allow them to participate in the decision-making process will result in greater motivation. satisfaction. and productivity. Orgunizurion feedbuck was defined as organizationally mediated performance information provided to the employee. Participution was defined as the degree to which the computer programmer is able to influence organizational decisions about that job. The hypotheses concerning organizational factors are: H4: The degree to which the computer programmer perceives that his or her supervisor initiates structure is related positively to that individual’s productivity and job satisfaction. H5: The degree to which the computer programmer perceives that his or her supervisor provides consideration is related positvely to that individual’s productivity and job satisfaction. H6: The degree to which the computer programmer perceives that he or she receives performance feedback from the organization is positively related to that individual’s productivity and job satisfaction. H7: The degree to which the computer programmer perceives that he or she is allowed to participate in organizational decisions affecting their job is positively related to that individual’s producitivity and job satisfaction.

4. Task Characteristics 3. Organization

Factors

Empirical evidence which the supervisor

indicates that the degree to initiates structure and pro-

Considerable research has focused on task characteristics that may be related to job satisfaction and job performance. For example, results of re-

search by Hackman and Lawler [ll] and by Hackman and Oldham [12] indicate that employees will be motivated if they perceive their jobs to be characterized by high skill variety, task significance. task identity, autonomy, and feedback. These were the task characteristics examined in this study. Specific definitions for the five core job dimensions are provided below [12]: 1. Skill variety. “The degree to which a job requires a variety of different activities in carrying out the work, which involves the use of a number of different skills and talents of the person.” 2. Tusk 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.” 3. Tusk 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.” 4. Autonom_y. “The degree to which the job provides substantial freedom, independence, and discretion to the individual in scheduling the work and in determining the procedures to be used in carrying it out.” 5. Feedback. “The degree to which carrying out the work activities required by the job results in the individual obtaining direct and clear information about the effectiveness of his or her performance.” James et al. [15] found employees’ perceptions of job variety, autonomy, importance, and challenge to be positively related to their level of job satisfaction. Sims et al. [22] found employees’ levels of job satisfaction and productivity to be negatively related to their perceptions of job pressure and the degree to which the job requires the employee to deal with others to complete task requirements, and positively related to the employees’ perception of task identity. Cougar and Zawacki [S] have conducted national surveys on more than 6000 persons in the computer field. Programmers and analysts were found to have very low social need strength (i.e., they have little need to interact with others). At the same time these individuals’ perception of their need to grow (i.e., growth need strength) is higher than any of the 500 job categories in the Hackman and Oldham study [12]. Cougar and Zawacki [5] suggest that these two characteristics of low social need strength and high growth need strength can be counterpro-

ductive. For example, computer professionals have a low need for meetings and they have not been effective in producing output from meetings yet effective interaction with system users is one of the most important factors involved in producing a complete and accurate set of information requirements [7,19]. The formal hypotheses are: HS: The degree to which the computer programmer rates his/her job high on the core dimensions (skill variety, task identity, task significance, autonomy, and feedback) is related positively to their levels of productivity and job satisfaction:

5. Dependent

Variables

Job satisfcrction is defined as the degree of compatibility between the individual’s needs. goals and values and the rewards provided by the job. Several studies [13,21,26,28], have used job satisfaction as a dependent variable. In most situations these researchers postulated the use of user job satisfaction as an objective in the systems design process. Bostrom [3] did utilize the satisfaction scales from the Job Diagnostic Survey [12] to measure the systems designers’ level of job satisfaction. The three dimensions that were included in this instrument were: 1. an overall measure of the degree to which the designer was happy in his/her work, 2. growth satisfaction (e.g., is the designer happy with his/her rate of personal growth and development), and 3. context satisfaction (which included the aspects of pay, security, social and supervisory variables). Bostrom did find some intercorrelations between the systems designer’s general job satisfaction measure and these individual dimensions (e.g. growth, r = 0.66; pay, r = 0.37; security, r = 0.40; social, r = 0.34; and supervision r = 0.40). In general, however, researchers have noted very low correlations between overall job satisfaction and sum of facet (i.e., context) satisfactions [1,28]. The Minnesota Satisfaction Questionnaire (MSQ) was used to measure context satisfactions and general overall satisfaction was measured by a single item which utilized a 5-point scale [27]. Progrummer producitivity has been discussed in many articles but few authors have addressed the topic scientifically. The problem of measuring the

productivity of programmers has not been solved to date: due.

to the nature

Programming therefore. I

in a concrete manner it may never be solved

of the programming

task

The instruments were either single questions. which utilized a Spoint Likert scale. or previously validated instruments. The use of IBM’s Programmer Aptitude Test (PAT) as a measure of mathematical aptitude may

itself.

is still more an art than a science: used the immediate supervisors’

evaluation on a Spoint scale as a perceived rogate measure of programmer productivity

he confusing, because the PAT test was initially set up to measure programming aptitude. The test

sur:

(l-

used

programmer is in the top IO’% in terms of productivity relative to the other programmers supervised: 2-programmer is in the next 20?,: 3programmer is average. in the middle 4OY”: 4programmer is poor but not the worst I’ve ever seen,

next

tom

10%).

20%;

and

5-programmer

programmers

such

variables

Table The

at each

of the two participating

The measurement in the

study

instruments are

presented

a pattern

choose

the

matical

ability

is a predictor

One of the if mathe-

of programmer

Variable

7. Results This study characteristics

set out to see if the several individual of computer programmers’, four

organizational factors and five task characteristics are related to computer programmer productivity and job satisfaction. The first step in the data

organi1.

Instruments Variable

Name

Type

Independent

1. Self-esteem

Instrument

Description

Self-esteem

measure

in the Jackson

question:

Independent

IBM

Programmer

Independent

House

and Desaler

(1974)

Instrumental

5. Consideration

Independent

House

and Deaaler

(1974)

Supportive

6. Feedback

Independent

Modified

version

7. Partvzipation

Independent

Hackman

and Oldhan

(1974)

Participation

(1974)

Job Diagmwic

3. Mathematical 4. Intiation

8. Core

Task

Aptitude

of Structure

Variables

9. Job Satisfaction 10. Productivity

How

Hackman

and Oldhan

Minnesota

Satisfaction

Dependent

Single other

programmers

Please

Leadership Leadership

and Oldhan

Questionnaire ratr

(Jackson

1976)

programmer?

Tebt

of the Hackman

Independent

Inventory

year5 have you been a computer

Aptitude

Dependent

question:

many

Perwnallty

Single

2. Experience

pro-

ductivity.

1 Measurement

next

it was never validated for this purpose appear as though I am begging the by using the test, but this is not the case

due to the lack of empirical validation. purposes of this study is to determine

used for the in Table

tests

question

collected via questionnaires that were by the researcher at one time and in

place

zations.

to find and

[ 141. It may

73%’ of

spondent was 31 years old and had 4.5 years tenure with the company. The definition of the computer programmer’s position for the purpose that performs the of this study was: “a person detailed program design, coding, testing. debugging, documentation, and implementation of commercially orientated information systems.”

one

student

of letters

3. The third had the student solve arithmetic word problems. This test is an adequate measure of mathematical aptitude (mathematical and logical ability), which was thought by IBM and others to be an indication of programming ability. but like man)

the computer programmers at the two firms (61 of 80 and 88 of 124 programmers). The average re-

Data were administered

the

a series

2. The second asked the student to do essentially the same thing but with figures not letters and:

is in the bot-

represented

parts:

asked

one:

were gathered from 149 computer in two midwestern corporations. The

149 computer

first

among

6. Procedure The data programmers

had three

1. The

Feedback

from

Agents

Scale Survey

and Single

this programmer’s

you have supervised.

Scale Scale

5.point

Item

productivity scale.

relatl\,e

to

Scale

Table

2

Pearson

Correlation

Coefficients

for Independent

Variables C0K

Math Self-esteem

Experience

Aptitude ~O.lXOl

Self-esteem

1 .oooo

0.2140

Experience

0.2140

1.oooo

Math

aptitude

Structure

Consideration

Feedhack

Participation

Task.\

0.1301

0.2109

0.281 I

0.2600

0.2140

0.2710

0.2941

0. I x47

0.2222

0.2104

0.31 I I

- 0.1801

0.2710

1 .oooo

- 0.2001

- 0.0471

Structure

0.1301

0.2941

~ 0.2001

1 .oooo

0.2147

0.0141

0.1511

0.2114

Consideration

0.2109

0.1847

~ 0.047 I

0.2147

1.0000

0.0303

~ 0.0714

0.1700

Feedhack

0.281 I

0.2222

0.1711

0.0141

0.303

I .ooo

0.1309

Participation

0.2600

0.2104

0.1511

~ 0.0714

0.1309

I .ooo

Core

0.2140

0.3111

0.2114

0.1700

0.1444

tasks

-0.2010 0.1411

analysis was to test the multicollinearity among the predictor variables by calculating the Pearson correlation coefficients. They appear in Table 2. Among the independent variables the highest intercorrelation was 0.311 between experience and the core task variables. The dependent variables of job satisfaction and productivity had a correlation coefficient of 0.371. F-ratios were then calculated using a general linear model procedure. These were used as a measure of the effect of each independent variable on computer programmer productivity and job satisfaction. A test of significance provides evidence within prescribed limits that there is actually a dependency between the independent or predictor variable and the dependent variable. The prescribed limits or levels of significance chosen in this study were N = 0.05 and (Y= 0.01 which indi-

Table

3

C’alculated

F-Values

Independent

Variable

Dependent Productivity

Variable Job Satisfxtion

1. Self-esteem

2.18 *

3.11 **

2. Experience

8.46 **

0.714

3. Math

0.918

0.646

Aptitude

4. Structure

4.809

5. Consideration

1.946

6. Feedback

4.92 **

8.14 **

7. Participation

5.18 **

6.11 **

8. Core

Task

**

1.01 3.14 *

Variables

a. Skill variety

0.314

3.01 **

b. Task

identify

0.410

0.401

c. Task

significance

1.14

2.29 *

d. Autonomy

3.18 **

3.00 *

e. Feedback

3.14 **

3.95 **

* = significant

at a = 0.05.

** = signifvzant

at a = 0.01

0.171 I

-0.2010

-0.1401

0.1411

0.1444 -0.1401 1.000

cates that we are either 99 percent or 95 percent sure that a dependency actually exists. The results are presented in Table 3.

8. Conclusions One of the most important findings is that the individual characteristic of experience positively affected productivity. This has long been suspected. but little empirical evidence existed to substantiate the claim that experienced programmers were more productive than inexperienced programmers. Other interesting questions for future study include: Does productivity keep rising with time?; Much? and How much is “enough”? Another individual characteristic. selfesteem, affected both productivity and job satisfaction. These results should be of benefit to the MIS practitioner in the recruitment and development of computer programmers. MIS managers should also note the organizational factors (structure, participation, consideration and feedback) that affected productivity and job satisfaction. Programmers have a definite need to participate in those decisions that affect them; they need feedback both for guidance and to satisfy their psychological needs with regard to performance. If they do not obtain this from their direct supervisors their productivity and satisfaction will suffer. In terms of the task characteristics, only one variable (autonomy) was positively related to productivity, and two variables (taks significance and autonomy) were positively related to the computer programmer’s level of job satisfaction. These findings appear to indicate that management must

214

Resecrrc~h

first indicate to the computer programming staff why each project is important and secondly, if they wish to have satisfied employees they must give their programmers independence in completing the job. This does not mean that programmers should not be structured and have feedback, but that they resent a supervisor constantly looking over their shoulder. Dickson and Powers [7] recommended future research to answer the following question: “What are the differences in managerial practices between successful MIS groups and those that are not successful?” Hopefully, this research has begun to provide part of the answer to that question.

;(ltlwls struc~ture, Proce.Kse,s, Behorror, ness

[ill

Publications,

Inc..

R.J.

Hackman.

Job

Characteristics.”

E.E.

Evaluation

of

Number

Reactions

to

Job

Design

Prqects.

Job

Diagnostic

of Jobs

and

“TechnIcal

of Admimstratlve

the

Report

Sciences.

Yale

1976.

Systems

Design

,wrrc.

II41

E.

York.

American

M.A..

Howell,

Aptitude

in

Process.”

Mumford

and

Elhevier, J.W.

for

tfunwrl

Harold

C’lwrc~e crtzl Crw-

Sackman

(eds.),

Ne\s

1975.

Vincent,

Computer

and

R.A.

Gay,

“Testing

P.yyc,holoK,‘ Re-

Programming.”

port.\, 1967, 20. 1251-1256. L.R. James.

A. Hartman.

M.W.

Between

for Work

Stebbins.

Psychologxal

Motivation.”

and

A.P. Jones.

Climate

and

a VIE

Penonrrel Psyhologr.

1977.

30. 229-254.

Models

and

A.P.

of Job

‘*Examination

Brief.

Hums

Satisfaction.”

Relo~wns, 1978.

R.P. Bagozzi. a Function

“Salesforce of

SItuational

Performance

Indiwdual

and

Difference,

31.

Journul of Morhrtrng

Factors,”

Satisfaction

ah

Interpersonal.

and

Reseurc~h. 1978.

15.517-531. ConJIm Hundrn,g urtd Power ,n the Redmgtz Promn: A F&I Stud,, Ir~rwtrgcrtrot~ of the Relutrorwhrp Between Monugen,mt InJormutron S~aw.\ Lkrs md Svsterns Mutntenunte Per.ronnel, unpublished doctoral disser-

[31 R.P. Bostrom,

tation,

University

(41 J.A. Breaugh, and Employee

of Minnesota.

Between

Performance.

Absenteeism.

“Acrrden~~~ of

tudes.

1978.

“Relationships

Munu~ement

Recruiting

Sources

and Work

Journcrl.

24.

11x1 E.E.

and

R.A. Versus

Zawacki. those

“Motivation

of their

Levels

Science

[61 D.R. Dalton, An Expanded

Munagument

and

and and

Todor,

Positive

Myths,

R.P.

“Turnover

Turned

Smith.

Systems

MISRC-WP-71-01.

and

Research University

Project

Reality,”

Illinois:

~241A.D.

Over:

Center

Manage-

Mangement

Working

of Minnesota.

Paper

und Hu-

Industry.”

Gibson.

(1980)

1954-204. Watson

Weir.

Ivancevich,

and

J.H.

Proceedings.

[271 D.J.

Weiss.

Between

Performance,

Carroll, Dallas.

VI

T.C. lished

Jobs

Leader

Reward

Absenteeism,

and

53

Computers for Bu.~rmx: A Bwness

Publications,

Inc.

of

for

Computer

Users

InternatIonal

Systems

a Method

Symposium

in

of Asseaaon Econom-

1974.

R.V. Dawis, for

letin

Oqanl-

A.B.

merit.”

“Minnesota

Donnelly,

“Antece-

Accrden~y of Mmu-

1969.

Effectiveness

ics of Isomatics,

and J.M.

and

“The

Support 13

McKemey

Huling, The Metrsureund Retrremmt, Chicago,

Inferences

Emphum.

Satisfying

the Per1982.

Studr of Computer Spe-

Journal of Occupotronul Psyrhologv.

Creating

Decision

Decision Sciences.

Motivation”,

and C.L.

McNally.

Satisfaction,”

Affecting

554-569.

Employee

1982.

In wz-X

and Subordinate

“Manual

“Factors [91 W. Fuerst, and P.H. Cheney. ceived Utilization of Computer-Based in the Oil

Behwior

on the

1970, 5, 277-298.

and

Kendall.

“Causal

Work

[261 M.

Minneapolis. Behavior

Cures.”

1980.

1971.

[RI M.G. Evans, “The Effects of Supervisory Path-Goal Relationship,” Orgunirutronul

mrrn Prrformmce,

Rand Szilagyi.

Monugerrul

InSeries.

and

1916. 19, 547-559.

L.M.

oJ Sotisfuctmn

~251H.J.

“MIS

Design

Associates.

gement Journal.

~231P.C.

Accrdem.v o/

Perspective.”

Powers,

Opinions

Research

Behavior

Reurrw, 4, 1980, 20-22.

Dickson

formation

W.D.

Causes

(221 H.P. Jr. Sims, A.D. Szllagyi. and D.R. dents of Work Related Expectancies.”

of

MIS

Employees.”

“Job

Job Sutr.$rc.t~on: il [211 E. Mumford, uolrsts, Lonson, Longmans. 1972.

Quarter~v. 1979, 3, 3, 41-56.

Systems

Lawler.

ment

Managers

Its

5X. 1980. 20-22.

Journal of Applied Ps,c~hologr. 54, 8, 1970. 305-312. C‘onc.epts for Manage1191 H.C. Jr. Lucas. Informutwn .~~..~t~tn.v ment, New York. McGraw-Hill, 1982. Manqmenr InJ~mmmon Srstems. C’hicago, [201 R. Jr. McLeod.

Atti1981,

Turnover:

Awuntv~~.

“Expectancies as Determinants of Perfor[I71 A.K. Korman, mance.” Journd oJ Appkd P.syc~hologv. 55. 1971. 218-222.

142-147. Cougar

“Labor

1161 c’. Jones. Munugenwzt

of Alternative

91-98.

[lOI J.L.

of Applred Ps,v holog,,. 197 I (

and E. Mumford. “The Design of Computer (131 B. Hedberg Systems: Man’s Vision of Man as an Integral Part of the

Model

[1] R. Aldag,

MN,

III. ” Employee

Jourwl

4. Department

Universit>.

References

[71 G.W. ment:

Bust-

55, 259-2X6.

“Relationships

[51 J.D. MIS

Texas:

Lawler,

and G.R. Oldham. “The I121 R.J. Hackman. Survey: An Instrument for the Diagnosis

v51

VI

Dallas.

1973.

the

G.W.

Minnesota

England.

and L.H.

Sat&action

Lofquist,

Questionnaire.

Studies m Vocutronul Rehuhrhtutwn.

1967. Bul-

22. Willoughby. Satisfaction doctoral

“Needs.

Interests.

of Data

Processing

dissertation.

University

Reinforcer Personnel.” of Minnesota.

Patterns unpub1970.