RESEARCH CONTfWUTlONS Management of Computing Computerized Performance Gordon B

RESEARCH CONTfWUTlONS Management of Computing Computerized Performance Gordon B - Description

Davis Editor Monitoring Systems Use and Abuse RH IRVING CA HIGGINS and FR SAFAYENI ABSTRACT An exploratory study of computerized performance monitoring and control systems reveals both positive and negative effects Responses of 50 clerical workers f ID: 25785 Download Pdf

117K - views

RESEARCH CONTfWUTlONS Management of Computing Computerized Performance Gordon B

Davis Editor Monitoring Systems Use and Abuse RH IRVING CA HIGGINS and FR SAFAYENI ABSTRACT An exploratory study of computerized performance monitoring and control systems reveals both positive and negative effects Responses of 50 clerical workers f

Similar presentations

Download Pdf

RESEARCH CONTfWUTlONS Management of Computing Computerized Performance Gordon B

Download Pdf - The PPT/PDF document "RESEARCH CONTfWUTlONS Management of Comp..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.

Presentation on theme: "RESEARCH CONTfWUTlONS Management of Computing Computerized Performance Gordon B"— Presentation transcript:

Page 1
RESEARCH CONTfW3UTlONS Management of Computing Computerized Performance Gordon B. Davis Editor Monitoring Systems: Use and Abuse R.H. IRVING, C.A. HIGGINS, and F.R. SAFAYENI ABSTRACT: An exploratory study of computerized performance monitoring and control systems reveals both positive and negative effects. R.esponses of 50 clerical workers from 2 organizations with computerized monitoring were compared to 94 individuals from organizations in similar jobs without computerized monitoring. The results indicate that computerized monitoring is associated with perceived increases in

office productivity, more accurate and complete assessment of workers’ performance, and higher levels of organizational control. Respondents indicate that managers overemphasize the importance of quantity and underemphasize the importance of quality in evaluating employee performance. Workers perceive increased stress, lower levels of satisfaction, and a decrease in the quality of their relationships with peers and management as a consequence of computerized monitoring. The relevance of existing models of performance monitoring is examined in light of these findings. 1. INTRODUCTION The

increasing computerization of office functions is providing management with new opportunities to monitor, assess, and control employee productivity. By designing software packages that automatically This research was supported by Contract No. ZER.36001-3.3362. Department of Communications. Canada. and the Natural Sciences and Engineering Research Council. Canada. 1986 ACM OOOl-0782/86/0800-0794 75a record counts of production, error rates, and time taken to process items, organizations are now obtaining accurate, quantitative data on worker performance. These computerized performance

monitoring and control systems (CPMCS) are the subject of contro- versy. Some authors [2, 201 see positive effects, such as increased productivity, a more accurate assess- ment of employee performance, and greater organi- zational control over workers. Others [7, 271, point to the potential negative effects on the employees themselves, such as increased stress and decreased job satisfaction. The purpose of this article is to provide empirical evidence on the perceived effects of computerized performance monitoring and control systems (CPMCS). To accomplish this, individuals in organi- zations

with computerized monitoring systems were compared to individuals in similar jobs but without computerized monitoring. Additionally, by analyzing retrospective data, changes that were perceived to occur as a consequence of introducing computerized monitoring are identified. 2. LITERATURE REVIEW The literature dealing with the effects of CPMCS is controversial. There is general agreement that the use of computerized systems will provide increased 794 Communications of the ACM August 1986 Volume 29 Number
Page 2
quantitative feedback to both superiors and subordi- nates, and that

management will have an increased capability to monitor and control performance. The disagreement as to the effects of CPMCS lies in the implications of these capabilities. Many authors are highly critical of computerized monitoring. Gregory and Nussbaum [7] point out that it will lead organizations to design office work along the lines of factory-style production lines. Myers [Zl] maintains that managers who seek con- trol through computerization will face problems re- sulting from a reduction in employee initiative and motivation. Licker [17] emphasizes the potential dif- ficulty of

obtaining reliable data on work habits within a computer mediated system. Cooley [3], al- though highly critical of computerized monitoring, suggests that these systems could be designed to en- hance the quality of work rather than demean it. Unions are strongly opposed to computerized monitoriog. They are concerned about issues of job security, individual privacy, and the manner in which management will interpret computerized per- formance data [6, 7, 121. An example of the potential problems of comput- erized monitoring is provided by Walton and Vittori [27] who describe the case of an

insurance firm that installed such a system. The result was fear and resentment among both operators and line managers and an increase in stress, to the point that productiv- ity decreased. Another example is provided by John- son et al. [ll] who studied 200 word processing cen- ters. They found that many organizations had stopped using computer-generated counts of produc- tion since they found them to be invalid, not cost effective, and degrading to employees. Some authors view the potential effects of CPMCS in a more positive light. For example, computerized systems will likely provide more

feedback to indi- vidual employees. This immediate feedback can act as a motivator [5, 8-10, 23, 261. Schick [25] ties the increased use of computerized performance monitor- ing to more objective ratings. Olson and Lucas [22] suggest that computerized office systems will allow more flexibility in work locations (remote work) and timing (flexible hours). Bair [2] points out the bene- fits to the organization in terms of objectivity and control. Other authors blame management for the dysfunc- tional effects of CPMCS. They attribute the negative consequences to the design, implementation, and op-

eration of the system [l, 13, 14, 18, 27, 281. These dysfunctional effects, which have aIso been ad- dressed in the literature on noncomputerized per- formance monitoring [16, 241, include overemphasis Research Contributions on quantitative performance measures and reduced worker satisfaction. Finally, Walton and Vittori [27] maintain that successful CPMCS must reflect a balance between a people-oriented design and a technically-oriented design. Presumably, a successful CPMCS will accu- rately reflect both the technical requirements inher- ent in the job and make provisions for the more

qualitative aspects of the work as well. 2.1 Research Questions We adopted an exploratory approach to the problem of identifying the effects of computerized perfor- mance monitoring systems. Our reasons were as fol- lows. First, the controversial nature of the literature creates a situation in which many conjectures are possible. Second, there are no explicit theories of computerized performance monitoring which could be used for generating specific hypotheses. Third, predicting the impacts of a given performance moni- toring system depends on the way the system is in- troduced and used. Seven

research questions are addressed. Although they are not predictive or directional, they do incor- porate many of the variables which are considered relevant in the literature. The first six examine pos- sible differences between CPMCS and non-CPMCS work environments. They are stated as follows. Are there any differences between individuals working in CPMCS and non-CPMCS environments in terms of their perceptions of: 1. 2. 3. 4. 5. 6. job satisfaction? the relative emphasis given to performance measures? the amount and usefulness of the feedback received? the closeness of the relationship

between in- dividual performance evaluations and the distribution of individual rewards? the closeness qf supervision? the level of participation in the performance evaluation process? The final research question addresses the conse- quences of introducing CPMCS. 7. What perceived changes were reported as consequence of moving from a noncomputer- ized to a computerized monitoring system? 3. METHODOLOGY 3.1 Data Sources Three insurance companies, a financial institution, and a government regulatory agency participated in August 1986 Volume 29 Number Communications of the ACM 795

Research Contributions Satisfaction (05) Not Satisfir?d Neutral 16 Satisfied 22 Total 46 Emphasis on R?rformance Measures (Q4) Quantity Underemphasized Appropriate 19 Overemphasized 31 Total 50 Quality Underemphasized 22 Appropriate 23 Overemphasized Total 50 Characteristics of Feedback Amount of Feedback (Q8) Not Enough i5: Enough Too Much Total 49 Usefulness of Feedback (QB) Not Useful Adequate 13 Useful 2; Total Importance of Evaluation for Rewards @lo) Not important Marginal Importance Important 41 Total 46 Closeness of Supervision (Q12) Not Close 16 Acceptable 10 Too Close 23 Total

49 Participation in Evaluation Process (Q13) Low Participation 30 Average High Participation Total 46 23 21 38 82 16 45 19 80 15 58 74 41 41 91 13 20 58 91 18 55 81 38 36 17 91 50 11 19 80 31 37 60 128 16 64 50 130 37 81 124 62 63 15 140 21 33 84 138 20 11 96 127 54 46 40 140 80 19 27 126 TABLE I. Direct Comparison-CPMCS/No-CPMCS 2.25 0.3247 23.79 15.3 0.001 0.0005 0.20 p = 0.9048 0.93 0.6281 8.08 0.0176 13.14 = 0.0014 0.84 = 0.6570 Note: All cells contain totals of individual responses. = significant at = 0.05 or better the study. Two of the insurance companies had ex- tensive computerized

monitoring systems which, at the time of data collection, had been in operation for less than two years. The other three organizations were chosen to provide a basis for comparison. All organizations had formal performance evalua- tion systems in which pay increments were tied to performance. No piece-rate pay schemes were in effect. The sample included 50 clerical workers from the two organizations with computerized monitoring and 94 clerical workers from the other three sites. Eighty-five percent of the respondents were female with an average of 4.5 years of experience in their current jobs.

In the three insurance companies the clerical work involved the assessment and payment of insurance claims. In the financial institution the clerks filled out stock transfer forms and pension claim forms. The government employees performed similar jobs involving government controlled licens- ing information. The employees in all organizations performed 796 Communications of fke ACM August 1986 Volume 29 Number
Page 4
Research Contributions short-cycle, production-oriented clerical work. This involved entering data in a highly structured format, with a limited amount of client

interaction. The sim- ilarity in jobs, combined with the possibility that computerized monitoring could have been imple- mented, provided the justification for using this lat- ter group as a basis of comparison. Quantitative data on individual performance were not collected due to the sensitive nature of the issue. In retrospect, the organizations seemed more con- cerned about the possibility that individual perfor- mance data would ‘upset’ their workers than were the workers themselves. 3.2 Instruments The primary data collection instrument was a 36 item questionnaire administered during a

semistruc- tured interview. It recorded an individual’s reactions to the existing performance monitoring system to- gether with his or her perceptions of the effects of computerized performance monitoring. Additionally, individuals were asked to give a reason for their response to a particular question. Three types of perceptual data were collected. The first provided a direct comparison of CPMCS and non-CPMCS respondents. For example, a question relating to feedback asked “How useful is the feed- back you receive about your performance?” The second type consisted of retrospective data from

CPMCS respondents only. A sample question asked “How does the use of computers to record perfor- mance data affect the fairness of the performance evaluation process?” The third type consisted of the comments provided by the respondents. Though all data are perceptual. they were collected either from different sources or using different approaches. 4. RESULTS A test of normality showed that the data were highly skewed. Consequently, nonparametric statistics (e.g., Chi-square, binomial) were used in the analysis. Data were grouped into three categories to counter the presence of sparse cells.

The three categories were: (1) scores 1, 2, and 3, (2) the midpoint (score level of satisfaction with the current monitoring sys- tem (Table II, Q5). This is surprising in light of the retrospective data which showed that significantly more respondents indicated a decrease in satisfac- tion (p 0.05). The comments associated with the scores were an- alyzed in the search for an explanation for these conflicting results. In the organizations with comput- erized monitoring, 26 negative comments were made compared to only 17 positive comments. In the or- ganizations without computerized monitoring

the re- sults were similar with 38 negative comments com- pared to 23 positive comments. However, it is the nature of the comments that is revealing. In the CPMCS organizations all of the negative comments concerned the computerized monitoring. They fo- cused on the overemphasis managers placed on counts (of work done), the lack of emphasis on qual- ity and customer relations, and the increased stress created by the monitoring. In the organizations with- TABLE II. Relative Comparison-Retrospective Data Only Overall Satisfaction Fairness Satisfaction Stress Talk ta Others Interest Co-workers

Supervisors Wl) (Q29) (Q30) (Q31) (~32) (033) (QW Management (Q35) More (CPMCS) (036) Emphasis on Performance Measures Quantitative (Q25) Quality (QW Quantity (Q27) Productivity (QW Characteristics of Feedback Completeness (Q14) Accuracy (Ql5) Feedback (Ql7) 19 22 48 22 12 I,$ &7* 37 13 50 13 11 26 50 15 27 50 23 17 48 12 32 49* 14 29 48= 18 16 39 41 49 t4 17 19 50 34 13 49 38 11 SO* 37 50 35 48 28 19 48 Importance of Evaluation for Rewards Importance (Q24) 2 25 20 47 Appropriate (016) 10 29 9 48 4). and (3) scores 5, 6, and 7. Results for the direct analysis, which compares CPMCS to non-CPMCS

respondents, are presented in Table I. The retrospective data, which are used to estimate the potential effects of introducing a CPMCS, are given in Table II. 4.1 Satisfaction with Performance Monitoring System A comparison of CPMCS to non-CPMCS respondents failed to reveal any significant difference in their Closeness of Supervision Control Closeness (Ql9) 1 35 13 49 (Q20) 11 16 21 48 Judge performance iansj 6 32 11 49 Participation in Evaluation Process Participation (Q23) 11 5 28 44 Note: All cells contain raw data. = decrease = increase .S = stay the same N = number in sample significant

binomial (D/I) at = 0.05 or batter August 1986 Volume 29 Number Communications of the ACM 797
Page 5
Research Contributions out computerized monitoring the negative comments did not have a central theme. There were com- plaints about supervisors, lack of standards, unfair evaluations, and the like. Consequently, we con- clude that although the level of satisfaction with the performance monitoring systems was the same in both groups of organizations, the source of dissatis- faction in CPMCS organizations was clearly attrib- uted to the computerized monitoring system. 4.2 Emphasis on

Performance Measures Respondents in CPMCS organizations indicated that significantly more emphasis was placed on quantity of work than did respondents in non-CPMCS organi- zations (Table II, Q4, 0.01). A significantly higher proportion of individuals in CPMCS organizations also felt that the quality of work was under- emphasized (p 0.01). The data in Table II (Ql6, 25-28) support these findings. A significant proportion (p 0.01) of indi- viduals (84 percent) said that there had been an in- crease in the emphasis placed on quantitative meas- ures of performance (Q25). The results also indicate

an increase in the per- ceived quantity of work produced (Q26) and in the perceived productivity of the department (Q28). Sur- prisingly, there was no consensus on whether the quality of work (Q26) had changed. The comments provide us with some insights on this lack of consensus. It appears that there are two classes of errors or quality defects. One class in- volves errors that can be caught by the system. Indi- viduals who referred to these errors said they had decreased simply because the errors were being caught by the computer. Another group of errors were those involving human judgment.

Individuals reporting these latter types of errors indicated that they had increased. For example, why spend a lot of time processing one insurance claim when you could be pushing through others to get your produc- tion count up. In summary, the presence of a CPMCS is associ- ated with a perceived increase in emphasis on quan- tity, with a corresponding perceived increase in pro- duction. There may also be a decrease in quality in situations involving human judgment, 4.3 Perceptions of Feedback by Workers A direct comparison of respondents in CPMCS envi- ronments to those in non-CPMCS

environments yielded no significant differences concerning either the amount or the usefulness of the verbal and writ- ten feedback received on their performance (Table I, Q8, 9). The retrospective data (Table II, Ql7) contra- dict this result since they show that a move to a CPMCS was associated with a perceived increase in the amount of feedback (p 0.05). Analysis of the comments produces an explana- tion for these contradictory results. Although there appears to have been an increase in the amount of written feedback associated with a move to a CPMCS; and although this feedback is perceived

to be more complete (p 0.05) and more accurate (JJ 0.05) than before, the amount of verbal feed- back did not change. Unfortunately, in the direct questions on feedback (Q8, 9), our subjects focused only on verbal feedback. In the retrospective questions ((214, 15, 17) they focused on the written reports. The tendency to respond in this manner appears to have been induced by the wording of the questions. 4.4 Importance of Evaluation for Rewards Although the majority of both CPMCS and non- CPMCS respondents view performance as closely re- lated to rewards (QlO), those in the CPMCS organi-

zations see the relationship as being significantly closer than do non-CPMCS respondents. The retro- spective data support these findings. A significant proportion of the respondents (p 0.05) stated that the availability of computer-generated counts re- sulted in the reward system being more closely tied to the evaluation system (Q24). The comments indicated that the availability of a measure of production gave the organization objec- tive grounds for basing rewards. Prior to the availa- bility of objective counts, the organization had to base its assessment on more subjective measures such as

supervisors’ opinions and self-reported counts. 4.5 Closeness of Supervision Individuals in CPMCS organizations felt they were more closely supervised (Ql2) than did individuals in non-CPMCS organizations (p 0.01). Although the retrospective data from the CPMCS group did not show a significant increase in the closeness of supervision (Q2O), the indicated change was in that direction. Related to closeness of supervision is the organiza- tion’s control over performance. The results relating to control over performance (see Table II), indicate an overwhelming (p 0.001) consensus among office

workers that the ability of the organization to con- trol its employees has increased as a consequence of the computerized monitoring. The comments imply that this is a direct consequence of management knowing more precisely the amount of work pro- duced. 798 Communications of the ACM August 1986 Volume 29 Number
Page 6
Research Contributions 4.6 Participation in the Evaluation Process There were no significant differences between CPMCS and non-CPMCS respondents in terms of the perceived level of employee participation in the evaluation process (Q13). Furthermore, the retro- spective

data (Q23) did not indicate any change in the level of participation. A likely explanation for this result is that the or- ganizations studied each have their own company- wide performance appraisal scheme. Consequently, the availability of computer-generated counts does not affect an employee’s participation in the process. With hindsight, a more useful question would have asked the degree to which an employee could influ- ence or change a performance rating as a conse- quence of the evaluation. 4.7 Perceived Changes as a Consequence of Computerized Monitoring A major component of this

research study involved the identification of perceived changes that occurred as a consequence of computerized monitoring of per- formance. The binomial sign test was used to iden- tify significant differences between individuals per- ceiving an increase and those perceiving a decrease on each of the XI variables. The responses in Table II indicate that computer- ized monitoring is perceived to have a substantial effect in 15 of the 20 variables studied. These in- cluded: 1. an increase in the completeness, accuracy, and appropriateness of information collected; 2. more feedback, a higher

level of organizational control, and a greater ability of the organization to judge performance; 3. an increase in the importance of quantitative measures in the evaluation and an increase in the importance of the evaluation itself; 4. higher quantity of work output and a corre- sponding increase in productivity; and 5. an increase in the level of stress, a decrease in satisfaction, and a decrease in the quality of relationships with peers, supervisors, and se- nior management. These results indicate that a move to a CPMCS is perceived to be associated with an enhanced capa- bility to control

the quantitative aspects of individ- ual performance. Concurrently, there appears to be a deterioration in the organization’s ability to control the qualitative aspects of performance. 4.8 Other Results A final question to all employees who had experi- ence with a computerized monitoring system was whether they would like a decrease, increase, or no change in the level of computerized monitoring. Of the 39 responses received, 18 (46 percent) wanted a decrease, (13 percent) wanted an increase, and 16 (41 percent) wanted the level to stay the same. Those wanting a decrease suggested that the

organi- zation should pay attention to other aspects of the job, not just the counts. Furthermore, they felt that the stress level was too high, leading to poor morale among workers. Those wanting an increase in com- puterized monitoring felt that the system was more fair since it recorded what everyone did. Several individuals commented that they only wanted an increase if the system considered all aspects of the job, not just the quantity of work produced. There was no consistent pattern to the comments of people satisfied with the current level of monitoring. Several respondents said they

were happy with the level of monitoring since it provided them with information on their performance. Others said that it was not the level of monitoring that was the problem, but how the numbers were used by management. One person commented that the counts were often used as “a big stick. 4.9 Limitations of the Study The study has several limitations. First, all compari- sons between the group with and the group without the CPMCS are meaningful to the extent that these two groups were similar to each other with respect to all other relevant dimensions. Another limitation relates to the type

of data col- lected. Retrospective data are not strong evidence for any before-after comparisons. They depend on the accuracy of the respondent’s memory and the way in which the memory is colored by the present state of the person. Furthermore, all data collected were perceptual. These are only valid in their own do- main. For example, lacking objective measures, it would be difficult to establish whether productivity had increased or not. A further constraint was a function of the organi- zations studied. They were reluctant to allow us to ask sensitive questions relating to individual

produc- tivity and also limited the amount of time per re- spondent. This latter limitation did not allow re- peated measures of the same questions for establish- ing interitem reliabilities. There were also some statistical problems. The variables in our study were highly correlated. When this happens the real type-1 error is inflated. Conse- quently, the possibility exists that some of the vari- ables which we have labelled “significant” may, in fact, not be significant. August 1986 Volume 29 Number Communications of the ACM 799
Page 7
Research Contributions The effects of the

above limitations are to some extent counteracted by a cross validation between the comparative data, the retrospective results, and by the opinions offered by the participants. To the extent that the three data sources produce consis- tent results, these findings are hard to refute. Never- theless, one should consider the study as exploratory and the findings as suggestive of potential causal relationships. 5. SUMMARY AND CONCLUSIONS Based on this exploratory study it appears that the introduction of computerized performance monitor- ing may result in a workplace that is less satisfying to

many employees. Among the possible contribut- ing factors are the increase in stress and organiza- tional control. Extensive use of numerical perfor- mance feedback seems to create a more competitive environment which may decrease the quality of so- cial relationships between peers and between super- visors and subordinates. A computerized performance monitoring system may provide an increase in the amount of quantita- tive feedback available to all levels of the organiza- tion. This quantitative information may underlie the general perception that these systems increase the fairness of the

performance evaluation system and increase the organization’s ability to judge perfor- mance. It seems that increasing the quantitative information available to management tends to mini- mize the emphasis given to subjective information in performance evaluations. Reducing the emphasis on subjective measures may improve performance evaluations only to the extent that job performance can realistically be measured using numerical counts. On the other hand, many aspects of an individual’s work may not lend themselves well to numerical measurement. One simple example is the thoroughness with which

a claims investigator processes a claim. Strict use of numerical counts leads to the undesirable attitude portrayed by the comment, “When in doubt, pay out! We found little evidence that workers were op- posed to a CPMCS in principle. Their responses were consistent with the views expressed by Kling and Scacchi [13] that, it is not the technology itself, but rather how it is used by management that deter- mines an individual’s reaction. Some of the findings may be viewed in the context of existing models of performance monitoring and control. For example, perceived increases in produc- tivity

may be explained by Lawler’s [15] expectancy model of extrinsic motivation since the outcome of performance was given as feedback to the workers as well as being tied to a reward structure. However, this finding may also be explained with a simpler stimulus-response model of behavior. It is possible to argue that the explicit measure- ments derived from the CPMCS have made the or- ganization more bureaucratic. Application of Mer- ton’s [19] model of dysfunctional consequences of bureaucracy may explain the rigidity of behavior in processing claims (e.g., focus on quantity). Similarly, emphasis

on quantitative measures and the corre- sponding goal displacement on the part of the workers may be explained by the concept of “overmeasurement [4]. The fact that different aspects of the results may be explained by different models and that these models do not explain all of these results (e.g., per- ceived decrease in quality of interpersonal relation- ships) suggests t.he need for a more comprehensive model of performance monitoring and control sys- tems. The issues that such a model should address include: the flow of performance related information in the organization, the process

whereby managers use this information, workers’ response to the pro- cess of measurement, the type of feedback and its subsequent influence on behavior, and potential changes in the underlying social structure of the or- ganization In addition to the conceptual challenge, develop- ment of this comprehensive model will depend on researchers’ ability to gather “sensitive” data on in- dividual performance, individual reactions to per- formance monitoring systems, and organizational monitoring and control decisions. Acknowledgments. We acknowledge the contribu- tions of Dr. Paul Hearty, who

initiated and super- vised our Department of Communications contract: and Dr. Thomas Whalen and Dr. John Syrotiuk, also of the Department of Communications, who made useful suggestions. Sue Munro, of the University of Waterloo, contributed substantially to the success of the study. We would also like to thank the reviewers for their excellent suggestions. REFERENCES 1. Ackoff. R.L. Management misinformation systems. Manage. Sci. 14, 4 (1967), 147-156. 2. Bair, J.H. Productivity assessment of office information systems tech- nology. In Emerging Office Systems. R. Landau. J. Bair, and J.

Seigman, Eds. Ablex Publishing, N.J., 1982, 159-166. 3. Cooley. M. Computerization-Taylor’s latest diguise. Economic and Industrial Democracy. Vol. 1.. Sage. London and Beverly Hills, 1960, 523-539. 800 Communications of the ACM August 1986 Volume 29 Number
Page 8
4. Etzioni, A. Modern organizations. In Orgnnizations and their Manag- ing. B.M. Gross, Ed., Free Press. New York, 1968. 5. Festinger, L. A theory of social comparison processes. Hum. R&f. 7, (1954), 117-140. 6. Gotlieb. CC., and Borodin. A. Social Issues in Computing. Academic Press, New York, 1973. 7. Gregory, J.. and

Nussbaum, K. Race against time: Automation in the office. Uffi Technol. People 1, (1982). 197-236. 8. Hackman, J.R., and Lawler, E.E., III, Employee reactions to job char- acteristics. 1. Appl. Psych. 55, (1971), 259-286. 9. Hackman, J.R., and Oldham, G.R. Work Redesign. Addison-Wesley, Reading, Mass., 1980. 10. Hackman, J.R. Oldham, G.R., Janson, R., and Purdy, K. A new strat- egy for job enrichment. In Readings in Sysfemafic Management of Human Resources. Addison-Wesley, Reading, Mass., 1979, 81-102. 11. Johnson, B.M., Taylor, J.C., Smith, D.R., and Cline, T.R. Innovation in word processing.

National Science Foundation Projecf IS181 1079. National Science Foundation, 1983. 12. Kasurak. P.C.. Tan, C., and Wolchuk, R., Management and the hu- man resource impact of the electronic office. Optimum. 13, 1 (1982). 57-68. 13. Kling, R., and Scacchi. W. Computing as social action: The social dynamics of computing in complex organizations. Adv. Compuf. 19, (1980), 249-327. 14. Kling, R., and Scacchi, W. The web of computing: Computer tech- nology as social organization. Adv. Compuf. 21, (1982), l-90. 15. Lawler, E.E., 111 Control systems in organizations. In Handbook of Industrial and

Organizational Psychology. M.D. Dunn&e, Ed., Rand McNally College Publ., Chicago, 1976. 1247-1291. 16. Lawler. E.E.. III. and Rhode, J.G. Dysfunctional effects of control systems. In Information and Confrol in Orgnniznfions. Goodyear, Palisades. Calif.. 1976, 84-94. 17. Licker, P.S. The effects of the automated office on professional per- formance. Working Paper 16-82. Faculty of Management, Univ. of Calgary, Alta., July 1982. 18. Markus, M.L. Power, politics, and MIS implementation. Commun. ACM 26,6 (June 1983), 430-444. 19. Merton. R.K. Bureaucratic structure and personality. Social Forces

18, (1940). 5filL568. 20. Mintzberg, H. The Sfrucfuring of Organizations. Prentice-Hall, Engle- wood Cliffs, N.J.. 1979. 21. 22. 23. 24. 25. 26. 27. 28. Myers, C., Ed. The lmpncf of Computers on Management. MIT Press, Cambridge. Mass., 1967. Olson, M.. and Lucas. HZ.. Jr. The impact of office automation on the organization: Some implications for research and practice. Com- mm. ACM 25311 (Nov. 1982), 838-847. Pettigrew, A. M. Information control as a power resource. Sociology 6,(1972), 188-204. Ridgway, V.F. Dysfunctional consequences of performance measure- ments. Admin. Sci. Q. 2. 2 (Sept.

1956), 240-247. Schick, M.E. The ‘refined’ performance evaluation monitoring sys- tem: Best of both worlds. Personnel J.. (Jan. 1980), 47-50. Vroom. V.H. Work and Mofivafion. Wiley, New York, 1964. Walton, R.E., and Vittori, W. New information technology: Organi- zational problem or opportunity? Off: Technol. People 1, (1983). 249-273. Zuboff, S. Computer mediated work-The emerging managerial challenge. Oft: Technol. People 1, (1982), 237-243. CR Categories and Subject Descriptors: H.4.1 [Information Systems Applications]: Office Automation General Terms: Human Factors, Management, Measurement

Additional Key Words and Phrases: performance monitoring and control systems, job design, implementation of computer systems Received 7/85; revised Z/86; accepted 4/86 Authors’ Present Addresses: R. H. Irving, Faculty of Administrative Studies, York University, 4700 Keele Street, North York, Ontario, Can- ada M3J lP3. C. A. Higgins, School of Business, University of Western Ontario. London, Ontario. Canada N6A 5B8. F. R. Safayeni, Department of Management Science, University of Waterloo, Waterloo, Ontario, Canada NZL 3Gl. Permission to copy without fee all or part of this material is granted

provided that the copies are not made or distributed for direct commer- cial advantage, the ACM copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Association for Computing Machinery. To copy otherwise, or to republish. requires a fee and/or specific permission. ACM Membership Service Call ACM’s toll-free (24 hour) 800 number, I-800-526-0359 ext. 92 In New Jersey: 1-800-932-0878 ext. 92 and use your credit card (Mastercard, Visa or Amex) to: RENEW: update your membership instantly JOIN: sign up for ACM membership on

the spot INQUIRE: request membership information quickly Take advantage ofthis convenient and speedy new service. (This service is available to members residing in the continental U.S. and Hawaii.) Association for Computing Machinery 11 West 42nd Street, New York, N.Y. 10036 (212) 869-7440 August 1986 Volume 29 Number Communications of the ACM 801