Jurison 1 Chapter 1 Perceived Value and Technology Adoption Across Four End User Groups Jaak Jurison Fordham University, USA This chapter explores the role end user perceptions in information technology adoption from the perspective of innovation diffusion theory. It is based on empirical data from a three-year longitudinal study of an information system implementation in an engineering organization. Data were collected on six different applications and their adoption by four categories of end users: engineering managers, project engineers, professionals, and secretaries. The data indicate a substantial variance across time, user categories, and applications in terms of adoption rates and perceptions of technology. The managerial implications of the results are that differentiated implementation strategies focused on specific end user categories are likely to be more successful than a single broadbrush strategy for all users. The results also suggest a framework for predicting technology adoption in the long run, based on initial adoption rates and user perceptions of technology. INTRODUCTION The introduction of technological innovations is fraught with many difficulties and uncertainties. Innovations based on information technology are the most challenging because they interact with end users in a variety of different ways and can lead to many different outcomes: some intended, some unintended. Problems with user acceptance of seemingly well-designed and sound information systems have been observed since the early days of Previously Published in the Journal of End User Computing, vol.12, no.4, Copyright 2000, Idea Group Publishing.
2 Perceived Value and Technology Adoption information technology (Lucas, 1975). Despite our growing body of knowledge, these problems continue to persist (Keen, 1981; Markus, 1983; Benjamin and Blunt, 1992; Markus and Benjamin, 1996). The issues of user acceptance have been explored from several different research streams, including organizational change and innovation diffusion theory. This chapter examines technology adoption from the latter perspective. The aim of this study is to examine how adoption rates and perceptions of technology vary across different applications and end user categories over time. The applications are part of an integrated office information system; the users are members of an engineering organization within a large high technology firm. The study explores the role of initial adoption rates and perceptions of technology in predicting the eventual outcome of the diffusion process. It differs from previous studies in three major ways. First, the study explores adoption of six different information system applications instead of a single innovation. Second, the locus of adoption is on groups of end users defined by their job categories. Past diffusion studies have been confined exclusively to either individual adopters or large organizational aggregates regardless of adopter job functions. Finally, the study is longitudinal, covering a three-year time period, in contrast with pre/post-test designs and retrospective studies. Knowing how different workers perceive information technology and how these perceptions affect their adoption rates is important because it helps managers design more effective implementation strategies and offers guidance for management intervention. This knowledge is also important for providers of various Internet-based products and services for developing effective marketing strategies. THEORETICAL BACKGROUND The findings from implementation research suggest that the most critical problems are not technical, but are related to organizational and implementation issues (Cheney and Dickson, 1982; Mankin et al., 1984; Markus and Keil, 1994). As a result, technical system characteristics have attracted less interest among information systems researchers. Recently Yetton et al. (1997) examined the influence of both system characteristics and implementation process on system success. They found that the characteristics of the innovation are critical for low task interdependence innovations, while the implementation process is more important for high task interdependence systems. Innovation diffusion theory recognizes that while the technical attributes of the innovation per se may be not significant, perceptions of technology do matter and are important factors influencing technology adoption.
Jurison 3 Rogers (1983) has synthesized over 1,500 studies into a theory of innovation diffusion. He defines diffusion as the process by which an innovation is communicated through certain channels over time among the members of a social system. He also defines innovation as an idea, practice, or object that is perceived as new by an individual or other unit of adoption. Although the innovations studied by Rogers do not include information systems, several researchers have found his framework useful for analyzing the adoption process of information systems (Perry and Kramer, 1979; Huff and Monroe, 1985; Brancheau and Wetherbe, 1990; Taylor et al., 1994; Prescott and Conger, 1995). Moore (1987) reviewed office automation and end-user computing literature and found the innovation diffusion model as an appropriate theoretical basis for the study and management of both types of information systems. Brancheau (1987) also considers the innovation diffusion model as the most suitable theoretical framework for this type of information systems applications because the model s focus on the individual adoption process is consistent with the degree of autonomy most knowledge workers have in carrying out their work. In a comprehensive review of empirical research in information technology diffusion, Fichman (1992) concludes that diffusion theory provides a useful perspective on one of the most persistent and challenging topics in the IT field, namely, how to improve technology assessment and implementation. Innovation theory has been found useful for studying information systems in a broader organizational context (Swanson, 1994). More recently, researchers are attempting to view technology adoption within the framework of organizational learning and change management (O Callahan, 1998). The primary concern of innovation diffusion research is how innovations are adopted and why some innovations are adopted at a faster or slower rate than others. As people evaluate an innovation, they decide whether to adopt or reject the innovation. Once adopted, the decision can also be reversed at a later time the decision to reject an innovation once it had been previously adopted is called discontinuance. The rate of adoption is the relative speed with which an innovation is adopted by members of the group. It is usually measured by the number or percentage of individuals who adopt an innovation in a specified time period. When the cumulative number of adopters is plotted over time, the result is generally an s-shaped curve. The slope of the s-curve represents the adoption rate that may vary from innovation to innovation. Rogers (1983) identifies five perceptual characteristics of innovations that help explain differences in adoption rates: relative advantage, compatibility, complexity, trialability, and observability. They are defined as follows:
4 Perceived Value and Technology Adoption Relative advantage the degree to which an innovation is perceived as better than the idea it supersedes. Compatibility the degree to which an innovation is perceived as being consistent with the existing values, past experiences, and needs of potential adopters. Complexity the degree to which an innovation is perceived as difficult to understand and use. Trialability the degree to which an innovation may be experimented with on a limited basis. Observability the degree to which the results of an innovation are visible to others. In general, the rate of adoption is positively related to perceived relative advantage, compatibility, trialability, and observability, and is negatively related to perceived complexity of the innovation. Rogers concludes, diffusion scholars have found relative advantage to be one of the best predictors of an innovation s rate of adoption. Tornatzky and Klein (1982) also found that relative advantage, along with compatibility and complexity, are the most significant factors in explaining relationships across a broad range of innovation types. Davis (1989), who studied information technology usefulness and ease of use, arrived at a major conclusion that perceived usefulness is a strong correlate of user acceptance. Iacovou et al. (1995) found a positive correlation between perceived benefits and adoption. These and other studies indicate a convergence of findings supporting the central role of perceived relative advantage in predicting user acceptance of information technology. RESEARCH SETTING AND METHODOLOGY The study was conducted in an engineering division of a large Fortune 100 firm. The division develops and produces a wide range of electronic assemblies and systems for marine applications. The organization is structured as a matrix, with several functional groups and a project engineering group. The study followed the implementation of an integrated office information system from the beginning through a three-year evaluation period. The system provided six different types of applications: word processing, e-mail, electronic filing, electronic calendar, decision support (spreadsheets), and project management. By concurrently providing the applications through a single integrated system it was possible to eliminate potential temporal effects and minimize the confounding influences of other organizational or implementation variables present in many empirical studies.
Jurison 5 The use and user perceptions of the different types of applications were measured using self-administered questionnaires. Measurements were made across four different categories of end users: managers, project engineers, professionals, and secretaries. Traditional studies of information systems have differentiated only between managerial and secretarial use. This study included other user categories because they represented a large percentage of workers in the organization and their work processes and information needs were significantly different from those of managers and secretaries. Professionals were engineers and administrative staff with no managerial responsibilities. Project engineers were defined as a separate management category because they tend to be more information and control oriented than functional and senior managers, whose work has more motivational and emotive content. The use of any of the applications was totally voluntary. The system was installed initially as a one-year pilot project with 56 users to evaluate the suitability of the system and to assess its cost-effectiveness prior to a commitment for long-term use. After six months of pilot operation, management decided to acquire the system on a permanent basis and extended it beyond the engineering organization. The total number of users reached 260 at the end of the third year of operation. Aside from adding users and the necessary processing capability to support them, the system remained relatively stable throughout the three-year study. The test group for the study consisted of 43 users: 9 managers, 10 project engineers, 15 secretaries, and 9 professionals. All managers and secretaries were highly experienced, (8 to 36 years of experience) but the experience level of project engineers and professionals varied over a wide range (1 to 25 years). Sixty-seven percent of the group had some prior computer experience. Almost all project engineers and professionals had used computers before, while only two thirds of the managers and about half of the secretaries had prior computer experience. This sample was maintained throughout the first year of the project. At the end of the third year, 20 randomly selected people were added to compensate for attrition through transfers, promotions and resignations. The final sample consisted of 13 secretaries, 5 professionals, 12 managers, and 8 project engineers. In total, 47 percent users were from the original test group. All added participants were carefully selected to assure that they were comparable to those in the original group and that they had access to the system for about three years. The careful control of the selection process was important for maintaining the integrity of the longitudinal design 1. All users attended training classes, followed by hands-on tutorial assignments. Help desk services were provided throughout the project.
6 Perceived Value and Technology Adoption Data were collected at the end of three months, six months, 12 months, and three years from the initial installation of the system. Users were asked to rate the frequency of their usage of each application on a scale of 1 to 5, 1 indicating non-usage and 5 indicating frequent usage. Only those who reported levels of 3 or above were considered to be adopters of the application in question. The perceived relative advantage was based on responses to the question: In case some applications have to be deleted, which ones would you miss the most? List in the order of priority, with the most important application first. Weights were assigned to the responses with a maximum weight of three given to application with the highest priority. At the time the project was initiated, we could not find an instrument with proven reliability and validity for relative advantage. Based on the feedback from a pretest with a group of 13 users, ranking was selected over an interval scale because it was understood better by the users and was therefore expected to yield more reliable results. We do not believe that this approach would significantly affect the validity of the study. All studies involving relative advantage tend to suffer from measurement problems because it is so broadly defined (Moore and Benbasat, 1982). Other studies (e.g. Davis, 1989, Iacovou et al., 1995) have also used broadly defined measures such as perceived usefulness and perceived benefits as a similar construct. RESULTS Table 1 shows the applications and the extent of their diffusions at different time periods. Diffusion levels were based on the percentage of adopters from the total number of users in each category. The data indicate a substantial variance in the adoption rates across time, applications, and user categories. Some applications were adopted very rapidly (e.g., e-mail and word processing) while certain other applications (e.g., project management and decision support) failed to be adopted even after three years. In some cases a peaking effect was observed in the first three months, suggesting that early adoption and subsequent discontinuance had taken place. On further examination, this effect appears to indicate experimentation and use on a trial basis prior to the adoption decision. In order to avoid potential distortion of data by the peaking effect, data from the next measurement period (six months) are used to represent the initial adoption rates because they provide a more reliable and consistent measure of true relative rates across applications. While the diffusion rates appear to be highly application dependent, there are also notable differences among various job categories. Professionals, as a group, appear to have the slowest diffusion rates in almost all applications, suggesting that they might be more likely to be late adopters. The
Table 1: Adoption statistics in terms of percentage of adopters Jurison 7
8 Perceived Value and Technology Adoption largest difference can be observed between professionals and secretaries, being statistically significant at p<.1 level. Managers in general appear to be adopting at a higher rate than project engineers are, although the difference was not statistically significant. Because the adoption rates are highly application dependent and sample sizes were relatively small, we must be careful about drawing conclusions regarding overall adoption rate variations among different groups of workers. Tables 2 and 3 show how various user groups perceived the relative advantages of different applications in the first year and at the end of the third year, respectively. The relationship between initial adoption rates (6 months) and corresponding perceptions of relative advantage was tested by bivariate correlation analysis, indicating positive correlation (r =.634, p <.001). The same data are shown graphically in Figure 1. It can be observed that the data points are clustered into two groups: a small group in high diffusion and high relative advantage region and a large group characterized by low to medium range of both diffusion and relative advantage. The first group represents those applications that have rapidly diffused throughout the workgroup. It includes electronic mail and word processing. Electronic mail was adopted at the fastest rate because it offered significantly better advantage over any existing communications media like telephone and company mail. Word processing was rapidly adopted as it was perceived to offer an advantage over other forms of document generation. Both are applications support activities Figure 1: Adoption vs. perceived relative advantage (1st year)
Jurison 9 Table 2: Perceived relative advantage of the applications in the first year * Table 3: Perceived relative advantage of the applications in the third year *
10 Perceived Value and Technology Adoption that are central to an office work. The second group reflects applications that are perceived of less value and therefore are adopted at a slower rate. Figure 2 shows the same variables measured three years after system installation. There is a stronger correlation between the two variables (r =.732, p <.001) than in the first year. A significant movement of data points from their original positions in Figure 1 can be observed. The group of applications that diffused rapidly at the beginning continued the diffusion process and in most cases reached the 100 percent level. The group with lower initial diffusion rates diverged into two groups: those that continued the diffusion process and those that discontinued (the percentage of adapters declined). Detailed examination of the data indicates that all discontinued applications were in the lower left-hand corner of Figure 1 below the dotted line, characterized by low adoption and low perceived relative advantage. Among the remaining applications, shown in the middle of Figure 1 between the two dotted lines, all but the electronic calendar applications continued to be adopted or remained effectively at the same level. The diffusion of electronic calendars is a noteworthy exception to the general pattern. After experiencing a reasonably rapid initial diffusion, suddenly many managers, professionals, and project engineers discontinued the use of electronic calendars. The unexpected change can be attributed to the introduction of a competing innovation: personal planning calendars/notebooks. Approximately eight months after the system was installed, these planning calendars were issued to managers and key employees. Being Figure 2: Adoption vs. perceived relative advantage (3rd year)
Jurison 11 portable, the calendars were more functional and flexible than computerbased calendars. Because they were issued to senior management first, they became also status symbols. For these reasons, many original users of electronic calendars switched to manual calendars. In effect, the manual calendars offered higher relative advantage than electronic calendars, causing the initial relative advantage of electronic calendars to drop in the third year, as shown in Tables 1 and 2. Another surprising observation is the low perceived relative advantage of project management software. One would expect this software to be of substantial value to managers and project engineers. A closer analysis reveals that there were problems with this particular application: it had limited output graphics and lacked sufficient flexibility for tracking project status. Initially it was used only on four projects, two of which discontinued its use within five months. While the remaining two projects continued using it with limited success, eventually management decided to phase it out and replace it later with another project management package from a different vendor. DISCUSSION A comparison of long-term outcomes with early adoption results allows us to draw some tentative conclusions. First, those applications that are perceived to offer high relative advantage from the start are adopted rapidly and are likely to reach organization-wide adoption without management intervention. Second, applications that are perceived to offer low value and are adopted slowly in the beginning are unlikely to gain acceptance in the long run. They are more likely to be discontinued over time. Finally, those applications with average perceived relative advantage and modest adoption rates are likely to continue to be adopted, but may never reach total adoption. These conclusions are consistent with the generalization developed by Rogers (1983) which states that the innovations with a high rate of adoption have a low rate of discontinuance. Managerial Implications The insights gained from this study have important implications for practicing managers. The wide variation in perceptions across different end users groups suggests that differentiated implementation strategies focused on specific job categories are likely to be more successful than a single broadbrush strategy for all users. The knowledge of initial adoption rates and perceived relative advantages can be useful for developing effective implementation strategies and management interventions. First, management has
12 Perceived Value and Technology Adoption Figure 3: Adoption vs. perceived relative advantage (1st year) to segregate applications into different categories of initial adoptions (see Figure 3). The first category, characterized by high adoption rates and high perceived relative advantage, stands out as a successful group which does not need much management support for continued adoption. The second category is distinguished by initial low adoption rate and low perceived relative advantage. Any application falling into this category requires prompt management attention. Management needs to decide whether to continue or discontinue the application for this group of end users. A decision to continue requires either a redesign of the application or work processes, additional training, or a combination of these. Applications in the third category are those with modest adoption rates and medium relative advantage. They can survive on their own, but can be made more successful with management intervention. Previous research has shown that management intervention has significant influence on technology adoption (Leonard-Barton and Deschamps, 1988). Management can get clues for intervention by examining the remaining innovation attributes in innovation diffusion theory, particularly compatibility and complexity. Therefore, when dealing with applications in this third category, management should examine them for compatibility with the needs, values, and job environments of the end users and look for opportunities to eliminate or reduce any incompatibilities. Implementation strategies should be focused on making the benefits (relative advantage) of the technology
Jurison 13 explicit through education and making it easy for users to master the system (reduce perceived complexity) through training and user support. Implications for Future Research The study also has several important implications for future research. It reinforces the view that innovation diffusion theory can be used effectively to analyze the process of integrating information technology and organizations. This approach is particularly appropriate for analyzing systems such as office information systems and decision support systems that cannot be mandated by management, but must be adopted on a voluntary basis by those who are to use them. This approach could also be used to explain the adoption of various types of new Internet-based products and services. The study also suggests that researchers should recognize the wide variations across different end user groups and applications in their research designs. This means that researchers must be careful about aggregating data and that future studies should focus on the factors that explain these differences. While the emphasis in this study was on the perception of relative advantage of technology on diffusion rates, it does not mean that other factors are unimportant. Future studies should take into consideration the remaining innovation attributes: compatibility, complexity, trialability, and observability. The fact that diffusion rates are highly application dependent suggests that compatibility may be a significant factor in the diffusion of information systems. Such future studies could benefit from a set of instruments developed by Moore and Benbasat (1991) that was not available at the time this study was initiated. Limitations of the Study As with all empirical studies, this study has some limitations that should be considered in interpreting the results. First, the sample size was relatively small, thus limiting the robustness of statistical tests. The size was constrained by the size of the test group that participated in the pretest survey prior to system implementation. Second, the addition of randomly selected users in the third year to compensate for subject mortality limited the panel design. Although the added respondents were selected carefully to assure that they were part of the organization and had access to the system during the same three-year time period, no statistical tests were made to determine if the sample was comparable to the original sample. Third, all results are based on single, self-reported measures and are therefore subject to potential bias. At the time the study was initiated, no
14 Perceived Value and Technology Adoption multiple item instrument for measuring perceived relative advantage with proven reliability and validity was found. The frequency of applications usage was based on self-reported measures because the system did not have the capability to monitor individual usage at the application level. The reliability and validity of self-reported vs. computer-monitored usage is a complex issue that has not been totally resolved at this time. While computer-monitored data are empirically more reliable measures of system usage, the self-reported data are not necessarily more valid (Rice, 1990). A recent study (Collopy, 1996) found that in self-assessment of system usage, light users tended to overestimate the use while heavy users underestimated theirs. Therefore, we have no reason to believe that self-reported usage data biased the results in any significant way. Finally, any research conducted in a single organization would have limited external validity and the results should always be treated cautiously. Whether the results are generalizable to other organizations and types of systems is an issue for future research. SUMMARY The findings from this study provide evidence of a variance across time, job types, and applications, in terms of adoption rates and perceptions of technology. This suggests that care must be taken when aggregating data and that research methods must focus on the factors that explain these variances. These findings equally suggest a need for highly differentiated organizational strategies for the same applications in different workgroups and for different applications within workgroups. The proposed framework can serve as a useful tool for developing these organizational strategies. Note: This is a significantly revised version of a paper published in the Proceedings of the 1993 ACM SIGCPR Conference. REFERENCES Benjamin, R.J. and Blunt, J. (1992). Critical IT issues: The next ten years. Sloan Management Review. 33(4), 7-19. Brancheau, J.C. (1987). The Diffusion of Information Technology: Testing and Extending Innovation Diffusion Theory in the Context of End-User Computing, unpublished Ph.D. dissertation, University of Minnesota, MN. Brancheau, J.C. and Wetherbe, J.C. (1990). The adoption of spreadsheet software: Testing innovation diffusion theory in the context of end-user computing. Information Systems Research. 1(2), 115-143.
Jurison 15 Cheney, P.H. and Dickson, G.W. (1982). Organizational characteristics and information systems: An exploratory investigation. Academy of Management Journal. 25(1), 170-184. Collopy, F. (1996). Biases in retrospective self-reports in time use: An empirical study of computer users. Management Science. 42(5), 758-767. Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology, MIS Quarterly. 13(3), 319-339. Fichman, R.G. (1992). Information technology diffusion: A review of empirical research. In: Proceedings of the Thirteenth International Conference on Information Systems. December 13-16, Dallas, TX, 195-206. Huff, S.L. and Monroe, M.C. (1985). Information technology assessment and adoption: A field study. MIS Quarterly. 9(4), 327-340. Iacovou, C.L., Benbasat, I., and Dexter, A.S. (1995). Electronic data Interchange and Small Organizations: Adoption and Impact of Technology. MIS Quarterly. 19(4), 465-485. Keen, P.G.W. (1981). Information systems and organizational change. Communications of ACM. 24(1), 24-33. Leonard-Barton, D. and Deschamps, I. (1988). Managerial influence in the implementation of new technology. Management Science. 34(10), 1252-1265. Lucas, H.C. Jr. (1975). Why Information Systems Fail. New York and London: Columbia University Press. Mankin, D., Bikson, T.K., and Gutek, B. (1984). Factors in successful implementation of computer-based office information systems: A review of literature with suggestions for OBM research. Journal of Organizational Behavior Management. 6(3/4), 1-20. Markus, M.L. (1983). Power, politics, and MIS implementation. Communications of ACM. 26(6), 430-444. Markus, M.L. and Benjamin, R.J. (1996). Change agentry the next information systems frontier. MIS Quarterly, 20(4), 385-407. Markus, M.L and Keil. M. (1994). If we build it, they will come: Designing information systems that people want to use. Sloan Management Review. 35(4), 11-25. Menard, S. (1991). Longitudinal Research. Newbury Park, CA: Sage Publications. Moore, G.C. (1987). End user computing and office automation: A diffusion of innovations perspective. INFOR. 25(3), 214-235. Moore, G.C. and Benbasat, I. (1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research. 2(3), 192-222.
16 Perceived Value and Technology Adoption O Callahan, R. (1998). Technology diffusion and organizational transformation: An integrative framework. In Information Systems Innovation and Diffusion: Issues and Directions, T.J. Larsen and E. McGuire (Eds.). Hershey, PA.: Idea Group Publishing. Perry, R.E. and Kramer, K.L. (1979). Technological Innovations in American Local Governments / The Case for Computing. New York: Pergamon Press. Prescott, M.B. and Conger, S.A. (1995). Information technology innovations: A classification by IT locus of impact and research approach. The DATA BASE for Advances in Information Systems. 26(2&3), 23-41. Rice, R.E. (1990). Computer-mediated communication system network data: Theoretical concerns and empirical examples. International Journal of Man-Machine Studies, 32, 627-647. Rogers, E.M. (1983). Diffusion of Innovations. 3rd Ed. New York: The Free Press. Swanson, B.E. (1994). Information systems innovation among organizations. Management Science. 40(5), 1069-1092. Taylor, J.R., Moore, E.G., and Amonsen, E.J. (1994). Profiling technology diffusion categories: Empirical test of two models. Journal of Business Research. 31, 155-162. Tornatzky, L.G. and Klein, K.J. (1982). Innovation characteristics and innovation adoption-implementation: A meta-analysis of findings. IEEE Transactions on Engineering Management. (EM 29-1), 28-45. Yetton, P., Sharma, R., and Southon, G. (1997). Successful innovation: The contingent contributions of innovation characteristics and implementation process. Proceedings of the Eighteenth International Conference on Information Systems. Atlanta, Georgia, 1-19. NOTES 1 Menard (1991) defines longitudinal research as research in which (a) data are collected for each item or variable for two or more distinct time periods; (b) subjects or cases analyzed are the same or at least comparable from one period to the next; and (c) the analysis involves some comparison of data between or among periods.