Impact of prototyping on user information satisfaction during the IS specification phase

Abstract

A study was carried out on an IS development project during the requirements specification of a DB-oriented information system. User Information Satisfaction (UIS) was measured three times during the phase: after the abstract specification of the system, after the interaction prototype, and after the interaction prototype, and after the system prototype. All representations described the identical information system. Assuming that the users' understanding of the system increased during the requirements specification phase, the study is a longitudinal, ex ante assessment of UIS and involves an exploratory analysis of the impact of uncertainty upon the assessment of UIS. The interaction prototype consistently gave the highest mean scores on UIS and the system prototype the lowest. This is consistent with earlier findings that protyping may create unrealistic expectations. The reliability scores of the UIS scales were high and generally increased during the three rounds of the specification process. The variance due to individual differences was considerably greater in the case of the system prototype than in those of the abstract specification and interaction prototype. This maybe explained by the uncertainty effect, which increases the proportion of neutral values.