+ All Categories
Home > Documents > Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Date post: 27-Oct-2014
Category:
Upload: sumbul-zehra
View: 44 times
Download: 2 times
Share this document with a friend
Popular Tags:
15
Restoring a Sense of Control during Implementation: How User Involvement Leads to System Acceptance Author(s): Ann-Marie K. Baronas and Meryl Reis Louis Source: MIS Quarterly, Vol. 12, No. 1 (Mar., 1988), pp. 111-124 Published by: Management Information Systems Research Center, University of Minnesota Stable URL: http://www.jstor.org/stable/248811 Accessed: 06/05/2010 01:47 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=misrc. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Management Information Systems Research Center, University of Minnesota is collaborating with JSTOR to digitize, preserve and extend access to MIS Quarterly. http://www.jstor.org
Transcript
Page 1: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Restoring a Sense of Control during Implementation: How User Involvement Leads to SystemAcceptanceAuthor(s): Ann-Marie K. Baronas and Meryl Reis LouisSource: MIS Quarterly, Vol. 12, No. 1 (Mar., 1988), pp. 111-124Published by: Management Information Systems Research Center, University of MinnesotaStable URL: http://www.jstor.org/stable/248811Accessed: 06/05/2010 01:47

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=misrc.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

Management Information Systems Research Center, University of Minnesota is collaborating with JSTOR todigitize, preserve and extend access to MIS Quarterly.

http://www.jstor.org

Page 2: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

Restoring a Sense of Control During Implementation: How User Involvement Leads to System Acceptance

By: Ann-Marie K. Baronas APM Inc. 810 7th Avenue New York, NY 10019

Meryl Reis Louis Boston University Center for Applied

Social Science Boston, MA 02215

Abstract User involvement has long been considered a critical component of effective system imple- mentation. However, the perspective has suf- fered from mixed results of empirical tests and the lack of a theoretical explanation for the rela- tionship (Ives and Olson, 1984; Baroudi, et al., 1986). Our purpose is to present a theoretically- grounded perspective to account for effects of involving users during implementation, and to provide an initial test of this perspective. We propose that: (1) system implementation repre- sents a threat to users' perceptions of control over their work and a period of transition during which users must cope with differences be- tween old and new work systems; (2) user in- volvement is effective because it restores or en- hances perceived control. Results of a field ex- periment designed as a preliminary test of this perspective are discussed.

Keywords: User involvement, implementation, system acceptance, system development

ACM Categories: H.4, K.4, K.6.1, K.6.2, K.6.M

Introduction The need for accurate and efficient processing of corporate information has resulted in the wide- spread introduction of computer-based informa- tion systems (CBIS). In turn, CBIS have encour- aged the collection of more information, the need for more capacity, and the development of more comprehensive CBISs. But CBIS adoption has not been synonymous with effectiveness; excell- ence in technical design alone has not insured system success. As a result, considerable effort has been directed toward implementing CBIS that users accept and use and for which they express satisfaction. Toward those ends, user involvement has been called upon to supplement quality of technical design in the quest for system success. However, there is significant evidence that little is known about how strategies of user involvement work - if and when they do. No theoretically grounded explanation for links with system success has been established, nor has a consistent meaning of user involvement been developed (Baroudi, et al., 1986; Ives and Olson, 1984).

In this article we take one step back. Rather than assuming that user involvement per se is the key, we examine the experience of the typical user immediately before, during, and after CBIS implementation. Borrowing from social psycholo- gy and organizational science, we assemble a theoretical perspective on the user's experience. We propose that system introduction is per- ceived by the user as a period of transition during which the normal level of personal control is threatened. We propose that activities that re- store a user's perception of personal control dur- ing system implementation will contribute to user acceptance and other aspects of system suc- cess (e.g., satisfaction with the new system, sys- tem usage). Several operational definitions of personal control are examined, including deci- sion choice, behavioral choice, and predictabil- ity. We then consider links between the concept of personal control and activities considered within user involvement approaches of the past. In CBIS implementation guided by such approaches, the key is to provide users with opportunities to restore or enhance their sense of personal control regardless of whether such activities constitute symbolic involvement or "true" user influence (Baroudi, et al., 1986).

As a first test of this perspective, we present the results of an exploratory field experiment. Treat- ment group members in the study were sub-

MIS Quarterly/March 1988 111

Page 3: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

jected to an implementation process in which activities intended to increase personal control were featured.' Differences between treatment and control groups in users' experiences of the implementation process and responses to the new system two weeks and two months after implementation are reported.

Implementation, Transition, and Personal Control The proliferation of CBIS is a fact of life in work organizations today. In addition, there has been a dispersal of that technology across organiza- tional roles and settings. Systems are no longer the exclusive domain of information system de- partments or external service bureaus. Data, software, and hardware are widely distributed. For example, they are situated in marketing de- partments, R&D labs, and sales offices. No lon- ger do the people who "command" the computer have systems as their primary activity or back- ground area of competence. Information system personnel who were involved with systems of the past have been supplemented with nontechnical personnel from all functional areas and all levels of the organization. These are people who need to provide system inputs and/or receive system outputs in order to do their jobs, but who may care little about the systems per se.

CBIS may be introduced among nontechnical personnel familiar with manual systems, or, as is becoming more and more frequent, among per- sonnel currently working with an automated sys- tem. According to Lucas (1981), workers using manual methods of information management must now "... change their behavior to make the system function" (p. 2). Even when implementa- tion involves a change from one CBIS to another, the user - especially the nontechnical user - is

1Although system implementation is typically defined as a process that begins with system design and in- cludes systems testing and installation, this study fo- cused on the installation phase of implementation. In this particular implementation setting, end users (both managers and clerical employees) were not involved in system design or modification. Rather, the identical CBIS (a standardized payroll and personnel manage- ment information system) was installed at each state government agency after design and modification had already occurred. The effort was coordinated by the agency eventually responsible for its implementa- tion (installation) in all state government agencies.

subjected to a period of learning, adjustment, un- certainty, and stress.

Many workers have resisted the introduction of computer technology and the implementation of different systems. In some cases, they have been unwilling to learn how to use the new sys- tem and continued to use the former system (Markus, 1979). Others have attempted to block implementation through sabotage (Dowling, 1979; Dickson, et al., 1967). By and large, end users have been competent at using the old sys- tem, unfamiliar with the new system, and risk averse. They have not sought opportunities to learn behavioral effort, have relied on "habits of mind" or schemas and scripts as guides in auto- matic cognitive and behavioral processing (Louis and Sutton, forthcoming). Generally, people re- sist participating in uncertain situations. Argyris (1971) has argued that workers may feel that the computer will usurp their own valued decision- making responsibilities. Ainsworth (1977) has identified lack of understanding, fear of job loss, and human nature as underlying causes of re- sistance. Similarly, Sanders and Birkin (1980) have argued that threat to security, reduction in social satisfaction, and reduction in self-esteem and reputation are reasons for resistance. Sales and Mirvis (1985) have suggested that workers' general willingness to accept change significant- ly affects their responses to computer system implementation.

User Involvement In light of the prevalence of user resistance dur- ing CBIS implementation, researchers and prac- titioners have experimented with corrective tech- niques. A favored method has been the involve- ment of users in the implementation process (DeBrabander and Edstrom, 1977; Lucas 1978). Prescriptions of user involvement during imple- mentation are not new.

Many reasons have been given for following this course. As outlined in a review by Ives and Olson (1984), user involvement is predicted to increase user acceptance by:

1. Developing realistic expectations about sys- tem capabilities (Gibson, 1977);

2. Providing an arena for bargaining and conflict resolution about design issues (Keen, 1981);

3. Leading to system ownership by users (Robey and Farrow, 1982);

4. Decreasing user resistance to change

112 MIS Quarterly/March 1988

Page 4: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

(Lucas, 1974); 5. Committing users to the system (Lucas,

1974; Markus, 1983).

User involvement is also predicted to improve system quality by (Ives and Olson, 1984):

1. Providing a more accurate and complete assessment of user information requirements (Norton and McFarlan, 1975; Robey and Far- row, 1982);

2. Providing expertise about the organization the system is to support (Lucas, 1974);

3. Avoiding development of unacceptable or un- important features (Robey and Farrow, 1982);

4. Improving user understanding of the system (Lucas, 1974; Robey and Farrow, 1982).

Although such views are intuitively appealing, they have received only mixed support. In par- ticular, the links between user involvement and system quality and acceptance have no theoret- ical grounds or consistent empirical results. In addition, as Ives and Olson (1984) have noted, "underlying cognitive and motivational charac- teristics" have been ignored (p. 590). Efforts here focus specifically on motivational and cognitive factors accounting for potential effects of user involvement on implementation out- comes. The next section discusses more specifi- cally the situation nontechnical users face during CBIS implementation.

The introduction of a CBIS typically represents a period of significant change for the employees who will be end users of the system. Seldom is this major change voluntary. Thus, as a new sys- tem is being implemented, users are undergoing involuntary transition, that is, a period of time during which significant changes occur in the aspects of one's work role.

Much has been written about transitions associ- ated with changes from one formal work role to another (e.g., hirings, promotions, transfers, layoffs). However, many of the same features distinguish transitions in which there is no

change in official role or title, but instead the orientation to the role is changed, as when re- sources, demands, or priorities change (Louis, 1980a). CBIS implementation represents such a transition. Change, contrast, and surprise have been identified as features of situations to be

negotiated, or coped with, during major transi- tions (Louis, 1980b). Change is defined as differ- ences in the objective features of old and new

systems (e.g., work procedures) and can be anti-

cipated. Contrasts are the differences in system features that turn out to be personally significant, which may differ among people going through the same transition. Surprise refers to significant differences between an individual's anticipations and later experiences. People make sense of and cope with these features of transitions by drawing on relevant past experience, by seeking guidance from others close at hand, and by tap- ping into standard operating procedures in the immediate work environment (Louis, 1980b).

We propose that user acceptance of new sys- tems will be facilitated as changes are realistical- ly anticipated (e.g., through input from know-

ledgeable sources), as contrasts are given free expression (e.g., through discussion among co- workers and between implementors and users), as surprises are minimized (e.g., through active previewing and reality testing) and assistance is provided in coping with them as they arise (e.g., through ready availability and coaching of sea- soned implementors). Implementation activities that assist users in making sense of and coping with changes, contrasts, and surprises should contribute to system success.

Perceived Control CBIS implementation is further considered to represent a transition situation in which users experience a threat of reduced control over their work. At a minimum, temporary decreases in perceived control are likely. In the perspective developed here, the links between user involve- ment and system success are proposed to stem from the effects of user involvement in reestab- lishing perceived control. A worker's desire for control is proposed to be the psychological mechanism underlying user involvement. The desire for control is triggered as the introduction of a CBIS threatens existing control. Techniques other than user involvement that enhance or re- store a user's perception of control are hypothe- sized to contribute to system success or accept- ance of the system by the user. Greenberger and Strasser (1986) have proposed and are in the process of testing relationships between indi- viduals' actions in response to uncertain events and their perceptions of control.

The concept of perceived control is borrowed from social psychology. (See Langer, 1983, for a comprehensive review.) The desire for control lies behind individuals' attempts to gain informa-

MIS Quarterly/March 1988 113

Page 5: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

tion from the environment (Heider, 1958). People are motivated to know the causes of their own and others' behaviors (Jones and Davis, 1965; Kelley, 1971), and the exercise of control en- hances self-perception and self-esteem through the association of control with mastery (White, 1959). Personal control has been defined pri- marily in terms of choice, predictability, responsi- bility, and ability to reduce or get relief from an unpleasant condition.

Decreases in personal control result in negative consequences for individuals and organizations such as increased stress (Averill, 1973), with- drawal (Langer and Rodin, 1976), sabotage (Allen and Greenberger, 1980), depression (Seligman, 1975), and reduced performance (Glass and Singer, 1972). Research has consis- tently demonstrated that personal control over unpleasant stimulation can reduce stress and lead to greater endurance (Pervin, 1963; Corah and Boffa, 1970; Straub, et al., 1971; Geer and Maisel, 1972).

Field studies have contributed to our under- standing in this area. For example, hospital pa- tients for whom the perception of control over stress had been induced required fewer pain re- lief drugs and sedatives and nurses reported that these patients exhibited less anxiety as well (Langer, et al., 1975). Langer and Saegert (1977) found that the unpleasantness of a high- density condition such as shopping at a crowded supermarket, could be significantly reduced by giving shoppers information about the expected effects of crowding before their exposure to it. In addition, treatment group subjects demonstrated increased performance on a complex task and better emotional response after the shopping experience.

Implementation and Control By extension, increasing users' perceptions of control during CBIS implementation would likely ease the stressfulness of their implementation experiences. To increase behavioral control, a variety of techniques can be employed. These include, but are not limited to, traditional mean- ings of user involvement. For instance, users (rather than system implementors or trainers) can be given control over the necessity and tim- ing of breaks from tasks such as entering data or learning how to use the new system. When users believe that they are the principal judges of the pacing and characteristics of implementation

tasks, stress and its effects during implementa- tion should be reduced as has been demon- strated in laboratory and field experiment set- tings. These techniques illustrate translation of the personal control concept in terms of the abil- ity to reduce or escape from unpleasant situa- tions. Sense of control can also be enhanced by increasing the predictability of the implementa- tion process, which can be achieved by giving workers more information about the course of implementation and by addressing concerns about how implementation of computer systems can be expected to impact their work. These techniques are integrated in the implementation strategy employed in this study.

Other research on decisional control demon- strates the use of choice as an operational defini- tion of personal control (Averill, 1973; Steiner, 1979). When subjects believe they have the power to choose, they perceive control over their environments. Stotland and Blumenthal (1964) found that people who were given the opportun- ity to choose the order in which they would take parts of an IQ test were less anxious before the test than no-choice subjects. Interestingly, the reduction in anxiety occurred as a result of ex- pected choice - even without immediate con- firmation of actually having choice. Other studies of choice have demonstrated a variety of per- formance enhancing effects (Perlmuter and Monty, 1979; Glass and Singer, 1972).

The effects of choice can also be applied to CBIS implementation, as strategies to increase choice represent yet another way to improve the imple- mentation process by increasing workers' sense of control. By involving end users in decisions relating to implementation, workers may become more invested in the success of the implementa- tion and more satisfied with the system through the social-psychological mechanism of per- ceived control. We propose that when em- ployees are given opportunities to enhance their perceived control during CBIS implementation- through choice, predictability, responsibility, and/ or ability to reduce or escape from stress - they will adapt to the change and accept the system more readily.

In summary, we propose a perspective on the experience of users undergoing CBIS imple- mentation. We propose that (1) computer imple- mentation is likely to be experienced by nontech- nical users as a period of transition during which users make sense of and cope with various dif- ferences between old and new systems and their

114 MIS Quarterly/March 1988

Page 6: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

anticipations of these differences; (2) computer implementation is likely to represent a threat to users' perceptions of control over work. The re- mainder of the paper presents results of a pre- liminary study that assesses the implementation strategy associated with threats to perceived control.

Research Method An exploratory field experiment was conducted to provide an initial test of our perspective on personal control. The study focused on the im- plementation of a computer-based payroll and personnel manacement information system in state government agencies. Groups of payroll and personnel clerks, equivalent in age, sex, and length of employment, were assigned to either control or experimental conditions. Subjects in the control group underwent an unmodified im- plementation process, while experimental group subjects were exposed to a modified imple- mentation process, designed to increase work- ers' sense of personal control. Questionnaires were administered to assess subjects' percep- tions of the system and its implementation on three occasions: ten weeks before the planned conversion, two weeks after the conversion, and two months after the conversion.

Hypotheses Based on the view that CBIS implementation among nontechnical users represents a threat to personal control, we hypothesize that interven- tions that enhance or restore personal control will have positive effects on users during and after implementation. More specific hypotheses are:

1. Treatment group subjects will report that im- plementation was more successful than will control group subjects.

2. Treatment group subjects will report that im- plementation was less stressful than will con- trol group subjects.

3. Treatment group subjects will consider the implementation schedule to have been more reasonable than will control group subjects.

4. Treatment group subjects will be more satis- fied with their jobs after CBIS implementation than will control group subjects.

5. Treatment group subjects will be more satis-

fied with the CBIS than will control group subjects.

Field site and situation The CBIS was a statewide personnel and payroll management information system which had been designed to standardize and automate per- sonnel and payroll information and generate re- ports for regulatory agencies. The system was a terminal access, menu-driven, online data man- agement and retrieval system, used to enter, up- date, and retrieve information on employees and positions. The system could not be used to per- form ad hoc analysis of payroll and personnel iniformation stored in agency databases. All aspects of system development had been com- pleted by the time of the first system introduction. The system had been developed without the par- ticipation of system users. Therefore, systems installation rather than full systems implementa- tion served as the situation in which the per- ceived control perspective was tested.

This CBIS represented the workers' first direct exposure to automated systems. Staff from the agency responsible for installing the computer system (also a state government agency) worked closely over a three month period with personnel and payroll staff from the agencies re- ceiving the system. Implementation steps were highly structured and standardized across sites, and involved interaction between implementa- tion personnel, agency staff (the users), and agency managers. The primary steps and their relative timing from date of cutover to the CBIS were as follows: (1) initial meeting introducing the system to agency managers (-12 weeks); (2) system demonstration for users (-10 weeks); (3) preparation of data conversion forms (-8 to -6 weeks); (4) three-day, off-site training for users (-4 weeks); (5) system testing and parallel pro- cessing (-4 weeks until cutover).

Subjects Subjects were 92 payroll and personnel em- ployees from 35 state government agencies in the northeast United States. Eighty-nine percent were female, 35 percent were under 30 years of age, and 34 percent were over 44 years of age. Forty-three percent had some college or technical training beyond high school, and 23 percent were college graduates; the remainder had high school level education. Forty-two per-

MIS Quarterly/March 1988 115

Page 7: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

cent were employed by "small" agencies (em- ploying fewer than 100 persons), 38 percent by "medium-size" agencies (100-999 employees) and 20 percent by agencies of 1000 or more employees. Thirty percent of the subjects had worked in their present jobs for less than a year, 23 percent for one to three years, and 47 percent for longer than three years. The 43 people who were employees of the 13 agencies converting to the computer system during the first three months of the study were assigned to the control group. The 49 subjects in the treatment group were employees of the 22 agencies experiencing implementation during the four subsequent months.

Treatment The relative deadlines for implementation steps and cutover dates were the same for all agencies in both groups. Changes to the standard imple- mentation plan were made to increase personal control among treatment group subjects through manipulation of choice (Perlmuter and Monty, 1979), responsibility (Langer and Rodin, 1976), and predictability (Glass and Singer, 1972). In- tervention strategies were repeated for each monthly control and experimental group. Cross- talking among personnel (subjects) from various groups would have been highly unlikely, as each agency was operating independent of any other agency. All members of the implementation team were trained in the control-enhancing imple- mentation techniques and were involved with both control and treatment group agencies (and users).

An initial feature of the treatment consisted of two modifications to the user demonstration ses- sions, which occurred 10 weeks prior to cutover. The purpose of the demonstration was to ac- quaint agency staff with the computer system. Control group subjects viewed a system demon- stration that emphasized the technical capabili- ties of the CBIS; in contrast, treatment group subjects participated in a demo session that in- cluded a presentation about the implementation process itself, highlighting what users could ex- pect. In addition, workers discussed in small groups the changes they thought implementation might bring, while implementation team mem- bers were present to answer questions and address concerns. These interventions were de- signed to enhance the predictability of the imple- mentation process for users which, in turn, was

hypothesized to increase their sense of control over their work.

Another technique involved increasing aware- ness among implementation staff of the control- threatening nature of the implementation pro- cess for users. In meetings with implementation staff, the experimenters discussed guidelines and techniques to be used by implementors for increasing users' sense of choice, predictability, and responsibility. For example, to increase pre- dictability, implementors were advised to let us- ers know in advance when on-site visits would occur, rather than simply arriving on-site un- announced. Similarly, to increase choice and re- sponsibility, implementors were encouraged to be clear and precise about "products" required of users, while at the same time allowing users to decide work deadlines and work methods.

A change was also made to the initial meeting at which the system was described to agency man- agers, who served as users' supervisors. In the control group, the meeting was conducted in lec- ture format; the implementation workplan was presented sequentially and the technical aspects of implementation were emphasized. In the treat- ment group, managers were briefed in a discus- sion format about possible user responses to the implementation process, including stress, fa- tigue, and frustration, as a means of increasing predictability. To increase the sense of choice and responsibility, managers selected dates for deadlines and events within narrow time periods. In this meeting, they were also asked to recall and to discuss their experiences of previous organizational changes. This indirect interven- tion into the implementation experience of sub- jects was considered necessary in order to effect the treatment in light of such organizational reali- ties as role relationships and influence patterns. It seems unlikely that efforts to effect workers' perceptions of control can succeed without in- tervening directly with end-users and indirectly through their supervisors.

Thus, the set of changes was designed to pro- vide a consistent treatment such that inputs and interactions concerning the CBIS across the work context would enhance users' sense of per- sonal control by increasing predictability, choice, and responsibility. In turn, increases in personal control are hypothesized to result in increased user satisfaction with both the implementation experience and the CBIS.

116 MIS Quarterly/March 1988

Page 8: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

Table 1. Descriptions of Variables

VARIABLE #ITEMS MEAN* S.D.

Q1 & Q3

1. User Information Satisfaction 9 5.17 1.05

2. Job Satisfaction 4 5.86 .89

Q2

3. Implementation Success 2 5.49 1.02

4. Implementation Stress 2 3.52 1.50

5. Schedule Reasonableness 1 5.13 1.73

6. Implementation Team Evaluation 6 6.16 .77

7. Managers' Attitudes Toward Implementation 2 5.32 1.28

Q3

8. Managers' Satisfaction With System 1 5.45 1.28

9. Retrospective User Information Satisfaction 9 4.96 1.26

* A seven-point scale was used. Higher values represent positive or favorable ratings.

Measures As stated earlier, effects of the modified imple- mentation process were assessed through the administration of three questionnaires to sub- jects: 10 weeks before the cutover date, two weeks after cutover, and two months after cut- over. In the first and third questionnaires, mea- sures of job satisfaction, as developed by Cam- mann, Lawler and Seashore (1975) were in- cluded, as were measures of user information satisfaction adapted from the standard Bailey and Pearson instrument (1983). A pretest in a sample agency revealed that the original Bailey

2 Although a generalized short form of the Bailey and Pearson instrument had been developed by Ives, et al. (1983), a pilot test revealed the desirability of adapting or "customizing" the measures of the field setting of this study. Customizing took the form of providing definitions of items and describing scales that were specifically meaningful for this setting and its participants. Baroudi has indicated in personal communications that he supports this decision and the procedure resulting in the measure of user in- formation satisfaction employed in this study.

and Pearson instrument was too cumbersome, time-consuming, and alienating. The number of items was considered excessive and the format unsatisfactory for this type of field setting. There- fore, a panel of judges was assembled to select a subset of items from the full Bailey and Pear- son measure. A Likert scale format was used. (For more detail on instrument development, see Baronas, 1986.)2

In the second questionnaire, users' evaluations of the implementation team, the reasonableness of the implementation schedule, and managers' attitudes toward implementation [from Bailey and Pearson (1983)] were included. Other mea- sures in the second questionnaire included us- ers' rating of implementation success and stress- fulness, which were adapted from Cammann, et al. (1975). In the third questionnaire, users' rat- ings of their managers' satisfaction with CBIS were also assessed. (Questionnaires were given directly to managers, however the sample size for managers was too small to permit statistically valid comparisons.) For a summary of measures,

MIS Quarterly/March 1988 117

Page 9: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

see Table 1. Appendix 1 contains a listing of items comprising each variable.3

Results

Pretest differences Chi-square analyses conducted on responses from the first questionnaire indicated that there were no significant differences between control and treatment groups in terms of sex, age, edu- cational level, agency size, or length of time in present position. Analysis of variance conducted on preimplementation perceptions of work dem- onstrated that prior to the intervention, there were no statistically significant differences be- tween control and treatment groups in the extent of job satisfaction or user information satisfaction with the existing payroll/personnel systems.

Manipulation checks Tests of the extent to which the modified imple- mentation was carried out were conducted. Targets for those tests included the end users and members of the implementation team who prepared users and their managers for introduc- tion of the system.

Two items in the second questionnaire were used as a manipulation check with users during implementation. Differences between control and treatment group means approached signifi- cance (p<.07) with respect to the extent to which users felt their opinions were listened to during implementation. No significant differences were found between control and treatment group sub- jects pertaining to their feelings about the extent to which managers encouraged participation during implementation. In hindsight, we might have done better with items that more directly assessed aspects of control such as users' sense of choice or predictability.

In addition, a series of meetings was held with system implementors who described their activi- ties to one another and to the researchers. These meetings were designed to check on and reinforce the treatment. Information was pro-

3 For copies of the questionnaires and more extensive information about the development of measures, see Baronas (1986).

vided through these meetings indicating that sys- tem implementors had taken specific actions to provide users with a sense of choice, predictabil- ity, and responsibility.

Treatment effects Table 2 reports users' ratings of the implementa- tion process collected through a questionnaire administered two weeks after cutover to CBIS. Although results were not significant at the .05 level, treatment group members gave higher rat- ings of implementation success, lending tenta- tive support to hypothesis one.

No differences were found between treatment and control groups in their perceptions of the stressfulness of implementation. Instead, results revealed that reported stress was high for both groups (see Tables 1 and 2) and that the treat- ment did not succeed in reducing stress among members of the treatment group. Thus, no sup- port was found for hypothesis two.

Treatment group members rated the imple- mentation schedule as significantly more reasonable than did subjects in the control group, supporting hypothesis three. This finding is especially revealing because the relative im- plementation schedules were identical. The only differences were that supervisors of treatment group subjects were given a sense of decision choice during the initial meeting's discussion of schedule, and treatment group subjects were given a description of the implementation process and timeframes during the system demonstration.

Users' ratings of implementation teams were sig- nificantly higher for the treatment group as were users' ratings of managers' attitudes toward im- plementation. These results provide both a check on the treatment and a measure of the effects of the treatment.

In the questionnaire administered two months af- ter conversion, users in the treatment group re- ported their managers to be significantly more satisfied with the computer system than did us- ers in the control group. This further indicates the success of the indirect portion of the treatment.

No significant differences between groups were found in job satisfaction after system imple- mentation; thus, hypothesis four was not sup- ported. Treatment group subjects reported signi- ficantly higher user information satisfaction than

118 MIS Quarterly/March 1988

Page 10: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

did users in the control group, lending support to hypothesis 5a (see Table 2).

An index of the difference between user informa- tion satisfaction with the old payroll/personnel systems and with the CBIS was computed by subtracting post- from preimplementation scores of system satisfaction. Results indicated that us- ers in the treatment group were significantly more satisfied with information from the compu- ter system than they had been with the former payroll/personnel systems, whereas users in the control group had given higher ratings to their former systems than they did to the new system.

In addition, subjects were asked in the third questionnaire to recall and provide retrospective evaluations of their former systems. As Table 2 indicates, the control group retrospective ratings

were significantly more positive than were treat- ment group ratings of the previous systems. Re- sults suggest that control group subjects re- mained attached to the old system, while treat- ment group subjects were already switching their attachments to the CBIS.

Discussion In this paper, we have developed an integrated theoretical perspective on the effects of user in- volvement during system implementation. We proposed that the process of implementing a new CBIS represents a situation of transition in which workers experience a threat to their sense of control over work, if not a direct loss of control. We argued that interventions which restore a worker's sense of control would reduce the

Table 2. Results of Analysis of Variance

CONTROL TREATMENT F

Questionnaire Two+

Ho 1: Implementation Success 5.28 5.76

2: Implementation Stress 3.34 3.66

3: Implementation Schedule 4.60 5.59 8.06***

Implementation Team 5.84 6.44 15.17*** Evaluations

Managers' Attitudes 4.99 5.60 5.15* Toward Implementation

Questionnaire Three+

Managers' Satisfaction 4.48 5.19 4.15*

4: Job Satisfaction 5.80 5.91

5: Satisfaction with the CBIS

5a: User Information 4.78 5.51 10.23** Satisfaction

5b: User Information 0.44a -0.60 9.05** Satisfactiont1 - User Information Satisfactiont3

5c: Retrospective 5.30 4.66 6.12** Satisfaction with Old System

* p<.05; ** p<1; = p<.001.

a A negative score indicates a preference for CBIS. + Questionnaire Two was administered two weeks after cutover;

Questionnaire Three was administered eight weeks after cutover.

MIS Quarterly/March 1988 119

Page 11: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

threatening quality of the implementation experi- ence and, as a result, would heighten a user's satisfaction with the new system. In this view, the active ingredient in user involvement is per- ceived control; user involvement is effective be- cause it restores or enhances perceived control.

Results of a preliminary test of the perspective were presented. In the study, we followed the lead of work in social psychology where per- ceived control is affected by manipulating predic- tability, responsibility, and choice. In the treat- ment developed here we intervened during an initial demonstration of the system to users, in the instructions to members of the implementa- tion team who would work directly with users and with their managers, and in briefings of users' managers. With users, the implementation pro- cess was explained, the normal fears and stress associated with such changes were addressed, and users were offered choices. With imple- mentation team members, the intervention in- creased their awareness of the control- threatening aspects of implementation, and iden- tified strategies for decreasing threats to user control. Managers were helped to anticipate us- ers' reactions and to recognize normal change processes; they were encouraged to avoid ex- pressing overtly negative attitudes about imple- mentation efforts.

This study contrasts with past studies of user involvement which have employed survey and case methods primarily (Ives and Olson, 1984). This study was a field experiment in which data were collected from control and treatment groups at three points in time. The study, though exploratory, focused on the installation of the same CBIS across 35 organizational settings, thus providing for experimental control and generalizability. In this way, problems inherent in "one-shot" data collection were avoided; analy- sis of subjects' experiences and perceptions over time was made possible, as was the com- parison of experiences between treatment and control groups.

As hypothesized, treatment group members were significantly more satisfied with the new system than were control group members. They were also more positive in their perceptions of interactions with system implementors and perceptions of the attitudes expressed by their managers. User information satisfaction after im- plementation differed between groups in terms of views of the new system and retrospective views

of the old system. Initially, groups did not differ in their views of the old system; after implementa- tion attitudes held by control group members about the old system were significantly more positive than were those of treatment group members. In addition, treatment group members preferred the new system to the old, while the reverse was true in the control group.

Several interpretations of these findings are plausible. A combination of increased resistance to the new system and entrenchment or height- ened preference for the old system among con- trol group members was observed. In addition, "unfreezing" on the part of the treatment group members resulted from their experiences during installation. However, such effects may be the result of factors other than the personal control intervention. In particular, control groups re- ceived the system earlier than did treatment groups. Although checks were made to insure that no changes in system features or docu- mentation had occurred in the interim, it is possi- ble that changes in more subtle aspects of the system, user interfaces, and/or implementation team member orientations may have taken place. Design of future experimental studies should ensure that there is no such time seg- regation of control and treatment group imple- mentation experiences.

Other limitations in the design and execution of the study should be noted. Future studies should focus on the full implementation process as the experimental context, rather than the installation phase alone. In addition, limitations in the instru- ments used in this exploratory study did not allow us to pinpoint the precise elements of the overall intervention which led to the observed results. For instance, users in the treatment group may have responded more favorably to the CBIS be- cause they interpreted gestures designed to in- crease their feelings of control as attempts to provide support and concern. To address such issues, future research could compare interven- tions that are oriented around perceived control versus experienced support.

Another issue that requires further study is the context of control. Rather than treating control as a global construct, changes are needed that focus on areas of control considered important by the user. Identification of areas of control spe- cifically threatened by system implementation could help target interventions for future research.

120 MIS Quarterly/March 1988

Page 12: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

Beyond these theoretical and empirical issues, there are practical implications of the study and the larger perspective on which it was based. Should future work confirm the role of personal control during implementation, system develop- ers may wish to anticipate and address issues of personal control triggered during implementa- tion. Implementors should make efforts to re- store or heighten users' perceptions of control. They may do so by facilitating experiences of predictability, responsibility, and choice for sys- tem users. In terms of predictability, for example, users could be made aware that it is normal and common for certain aspects of control to be threatened or reduced during implementation. In terms of behavioral choice, users could be en- couraged to exercise control in other relevant areas not threatened. In essence, this strategy helps to differentiate among control domains by separating those in which the user's control re- mains or is heightened from those in which loss is likely.

In addition, opportunities for various decision choices could be developed, as was demons- trated in the intervention used in this study. Choices (within predetermined ranges) for spe- cific scheduling of deadlines, meetings, and even breaks could be left to users. Further, as system developers are able to get users to accept responsibility for particular implementa- tion tasks, they are likely to enhance users' perceptions of control.

Overall then, a comprehensive approach to re- ducing the control-threatening nature of system implementation makes more sense than does a piecemeal approach. System developers and implementors should make an effort to (1) give users a complete and accurate picture in ad- vance of their likely experiences during and after implementation - make it predictable; (2) find areas in which users can make meaningful deci- sions throughout the process - provide choice; (3) get users to "sign up," to be accountable for results on tasks necessary to the implementation effort - engender a sense of responsibility. Together, success in facilitating users' experi- ences of predictability, choice, and responsibility during CBIS implementation should be associ- ated with a heightened sense of personal con- trol, which is naturally threatened during system implementation.

In this perspective, we have seen that what mat- ters are users' perceptions of control, rather than external or objective assessments. By extension,

developing opportunities for users to exercise or restore control is not enough. They are no guarantee of success. More important than the actual changes implementors might make are their skills at communicating them to users, and linking them into users' experiences. There is a final implication of the perspective we have pro- posed. Efforts to provide for extensive user in- volvement during systems development may have less effect on users' acceptance of and satisfaction with systems than less elaborate in- terventions targeted specifically at restoring us- ers' perceptions of control.

References Ainsworth, W. "The Primacy of the User (Part I),"

Infosystems, April 1977, pp. 46-48. Allen, V.L. and Greenberger, D.B. "Destruction

and Perceived Control," in Advances in En- vironmental Psychology (Vol. 2), A. Baum and J.C. Singer (eds.), Erlbaum, Hillsdale, NJ, 1980, pp. 5-109.

Argyris, C. "Management Information Systems: The Challenge to Rationality and Emotional- ity," Management Science, 17:6, 1971, pp. B- 275-B-292.

Averill, J.K. "Personal Control Over Aversive Stimuli and Its Relationship to Stress," Psychological Bulletin, 80, 1973, pp. 286-303.

Bailey, J.E. and Pearson, S.W. "Development of a Tool for Measuring and Analyzing Computer User Satisfaction," Management Science, 29:6, 1983, pp. 519-529.

Baronas, A.K. "Psychological Reactance, Per- ceived Control and Computer System Imple- mentation: A Social Experiment," doctoral dissertation, Boston University, 1986, Uni- versity Microfilms International.

Baroudi, J.J., Olson, M.H., and Ives, B. "An Empirical Study of the Impact of User Imple- mentation on System Usage and Information Satisfaction," Communications of the ACM, 29:3, 1986, pp. 232-238.

Cammann, C., Lawler, E., and Seashore, S. Michigan Organizational Assessment Pack- age, Institute for Survey Research, Ann Arbor, MI, 1975.

Corah, N.L. and Boffa J. "Perceived Control, Self-Observation, and Response to Aversive Stimulation," Journal of Personality and Social Psychology, 16, 1970, pp. 1-14.

DeBrabander, B. and Edstrom, A. "Successful Information Systems Development Projects,"

MIS Quarterly/March 1988 121

Page 13: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

Management Science, 24:2, 1977, pp. 191- 199.

Dickson, G.W., Simmons, J.K., and Anderson, J.C. "Behavioral Reactions to the Introduction of a Management Information System at the U.S. Post Office: Some Empirical Observa- tions, working paper, Carlson School of Man- agement, University of Minnesota, 1967.

Dickson, G.W. and Simmons, J.K. "The Be- havioral Side of MIS," Business Horizons, 13:4, 1970, pp. 1-13.

Dowling, A.F., Jr. "Hospital Staff Interference with Medical Computer System Implementa- tion: An Exploratory Analysis," CICR No. 49, Center for Information Systems Research, MIT, Cambridge, MA, 1979.

Geer, J.H. and Maisel, E. "Evaluating the Effects of the Prediction-Control Confound," Journal of Personality and Social Psychology, 23, 1972, pp. 314-319.

Gibson, H.L. "Determining User Involvement," Journal of Systems Management, August 1977, pp. 20-22.

Glass, D.C. and Singer, J. Urban Stress, Academic Press, New York, NY, 1972.

Greenberger, D.B. and Strasser, S. "Develop- ment and Application of a Model of Personal Control in Organizations," Academy of Man- agement Review, 11:1, 1986, pp. 164-177.

Heider, F. The Psychology of Interpersonal Re- lations, Harcourt Press, New York, NY, 1958.

Ives, B., Olson, M.H., and Baroudi, J.J. "The Measurement of User Information Satisfac- tion," Communications of the ACM, 26:10, 1983, pp. 785-793.

Ives, B. and Olson, M.H. "User Involvement and MIS Success: A Review of Research," Man- agement Science, 30:5, 1984, pp. 586-603.

Jones, E.E. and Davis, K.E. "From Acts to Dis- positions: The Attribution Process in Person Perception, in Advances in Experimental So- cial Psychology (Vol. 2), L. Berkowitz (ed.), Academic Press, New York, NY, 1965.

Keen, P.G.W. "Information Systems and Organi- zational Change," Communications of the ACM, 24, 1981, pp. 24-33.

Kelley, H.H. Attribution in Social Interaction, General Learning Press, Morristown, NJ, 1971.

Langer, E.J. The Psychology of Control, Sage Publications, Beverly Hills, CA, 1983.

Langer, E.J., Janis, I.L., and Wolfer, J.A. "Re- ductions of Stress in Surgical Patients," in Advances in Environmental Psychology, Vol. 2; Applications to Personal Control, A. Baum,

and E. Singer (eds.), Lawrence Erlbaum Associates, Hillsdale, NJ, 1975.

Langer, E.J. and Saegert, S. "Crowding and Cognitive Control," Journal of Personality and Social Psychology, 35, 1977, pp. 275-282.

Langer, E.J. and Rodin, J. "The Effects of Choice and Enhanced Personal Responsibility for the Aged: A Field Experiment in an Institu- tional Setting," Journal of Personality and So- cial Psychology, 34, 1976, pp. 191-198.

Louis, M.R. "Career Transitions: Varieties and Commonalities," Academy of Management Review, 5:3, 1980a, pp. 329-340.

Louis, M.R. "Surprise and Sense Making: What Newcomers Experience in Entering Unfamiliar Organizational Settings," Administrative Sci- ence Quarterly, 25:2, 1980b, pp. 226-251.

Louis, M.R. and Sutton, R.I. "Switching Cogni- tive Gears: From Habits of Mind to Active Thinking," Organizational Behavior and Hu- man Decision Processes, forthcoming.

Lucas, H.C., Jr. "The Evolution of an Information System: From Key-Man to Every Person," Sloan Management Review, Winter, 1978.

Lucas, H.C., Jr. Implementation: The Key to Successful Information Systems, Columbia University Press, New York, NY, 1981.

Lucas, H.C., Jr. Toward Creative Systems De- sign," Columbia University Press, New York, NY, 1974.

Markus, M.L. "Power Politics and MIS Imple- mentation," Communications of the ACM, 26:6, 1983, pp. 430-444.

Markus, M.L. "Some Neglected Outcomes of Organizational Use of Computer Technolo- gy - and Their Implications for Systems De- signers," Proceedings of the National Compu- ter Conference, 1979, pp. 397-402.

Norton, D. and McFarlan, F.W. The Information Systems Handbook, Dow Jones-Irwin, Home- wood, IL, 1975, pp. 517-528.

Perlmuter, L.C. and Monty, R.A. (eds.). Choice and Perceived Control, Lawrence Erlbaum Associates, New York, NY, 1979.

Pervin, L.A. "The Need to Predict and Control Under Conditions of Threat," Journal of Psychology, 31, 1963, pp. 570-587.

Robey, D. and Farrow, D. "User Involvement in Information System Development: A Conflict Model and Empirical Test," Management Sci- ence, 28:1, 1982, pp. 73-85.

Sales, A. and Mirvis, P. "The Impact of New Technology on People in Organizations: A Re- view of the Current Literature," Boston Uni- versity Working Paper, Center for Applied So- cial Science, 1985.

122 MIS Quarterly/March 1988

Page 14: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

Sanders, D.H. and Birkin, S.J. Computers and Management in a Changing Society, McGraw-Hill, New York, NY, 1980.

Seligman, H.E.P. Helplessness, W.H. Freeman, San Francisco, CA, 1975.

Steiner, I.D. "Three Kinds of Reported Choice," in Choice and Perceived Control, L.C. Perl- muter and R.A. Monty (eds.), Lawrence Erlbaum Associates, New York, NY, 1979.

Stotland, E. and Blumenthal, A. "The Reduction of Anxiety as a Result of the Expectation of Making a Choice," Canadian Journal of Psychology, 18, 1964, pp. 139-145.

Straub, E., Tursky, B., and Schwartz, G.E. "Self Control and Predictability: Their Effects on Reactions to Aversive Stimulation," Journal of Personality and Social Psychology, 18, 1971, pp. 157-162.

White, R.W. "Motivation Reconsidered: The Concept of Competence," Psychological Re- view, 66, 1959, pp. 297-333.

About the Authors Ann-Marie Baronas is an Associate with APM, Inc., a management consulting firm based in New York City. Prior to joining APM Dr. Baronas

was a consultant with the Commonwealth of Massachusetts' Office of Management Informa- tion Systems, where she conducted research on MIS implementation and computer use. She re- ceived her Ph.D. in social psychology from Bos- ton University in 1986.

Meryl Reis Louis is an Associate Professor at Boston University's School of Management and a research associate at the Center for Applied Social Science. Before returning to UCLA's Graduate School of Managtment for a docto- rate, she served on the consulting staff of Arthur Andersen & Co. and worked as a counselor in a community mental health center. Professor Louis has been a member of the editorial boards of Administrative Science Quarterly, the Academy of Management Review, the Organi- zational Behavior Teaching Review, Organiza- tional Dynamics, and Consultation. Her research interests have centered on cognitive processes in work settings, career and other work transi- tions, workplace cultures, information systems in strategic human resource planning, and the sociology of social science. For the past 10 years, she has been studying "life after MBA school" with a panel of graduates from four ma- jor MBA programs.

MIS Quarterly/March 1988 123

Page 15: Ann Mari k Baronas Meryl Reis Restoring Sense of Control

Implementation Control

Appendix I Listing of Individual Items

Comprising Measures Used in the Study Variable 1. User Information Satisfaction:

* Accuracy - The correctness of the informa- tion contained in the system.

* Currency - The age, timeliness, or "up-to- dateness" of the information contained in the system.

* Completeness - The completeness, or com- prehensiveness, of the information contained in the system.

* Flexibility - The capability of the system to change or adjust in response to new condi- tions, demands, or circumstances.

*The degree of understanding I have about the system.

* The level of confidence I have about the in- formation contained in the system.

* The level of control I feel over the system.

* My level of familiarity with the system.

* My general level of satisfaction with the sys- tem.

Variable 2. Job Satisfaction:

* I get a personal feeling of satisfaction doing my job.

* I don't care what happens to the organization as long as I get my paycheck.

* In general, I like working here.

* All in all, I am satisfied with my job.

Variable 3. Implementation Success:

* System implementation was a success at my agency.

* All in all, I am satisfied with the way imple- mentation was handled at my agency.

Variable 4. Implementation Stress:

* Implementation was a stressful process.

* Too many demands were placed on staff like me during implementation.

Variable 5. Schedule Reasonableness:

* Schedule of implementation - The time- table for converting your agency to the system.

Variable 6. Implementation Team Evaluation:

* Technical competence of the implementation staff.

* Payroll/personnel knowledge of the imple- mentation staff.

* The implementation team's relationship with the staff of your agency.

* Attitude of the implementation staff.

* Concern for users of the implementation staff.

* Implementation process - The manner in which the computer system was introduced into your agency by the implementation staff.

Variable 7. Managers' Attitudes toward Imple- mentation:

* My manager did not agree with the imple- mentation staff about how to implement the system.

* My manager was unhappy about our agency going on the system.

Variable 8. Managers' Satisfaction With The System:

* My manager is satisfied with the computer system.

124 MIS Quarterly/March 1988


Recommended