and organizational factors affecting system implementation (Ives and Olson, in press; RobeyandFarrow, 1982; Fuerstand Cheney, 1982; Lucas 1976).However,littleresearchhas captured the dynamics of user involvement during system development Consequently, we 1mow little about such activities and events as user and designer interaction, top management pressures, political activities, or the role of personnel turnover. To team more about these dynamics, longitudinal research needs to be conducted throughout the system development process. Longitudinal data collection serves two important research objectives: improved measurement validity, and alternative interpretations of events. Events may be described in a more valid manner when data is collected as events occur. This guarantees greater independence of the construct "user involvement" from the construct "system success." In reconstruction studies, respondents' current experience with a system may influence their perceptions of involvement Second, longitudinal study makes it easier to collect alternative interpretations of events in the development process. Activities that are ostensibly undertaken for rational reasons (to deliver a better system) may also serve the political interests of the actors (Robey and Markus, 1984). Longitudinal methods permit respondents' interpretations of these events to be collected as they occur. Thmugh longitudinal research methods, the events of development processes may be described and then interpreted in two ways. One interpretation stresses the rational objectives and methods in development The second interpretation of the same events identifies political themes. These two perspectives coexist and complement one anothen