System Design Features and Repeated Use of Electronic Data Exchanges

Oftentimes researchers may not only generalize across a population, but may also extrapolate research findings across time. While either assumption can introduce difficulties, generalizing results in one time frame to another time frame may be especially perilous. We study a data exchange, and find that interventions designed to improve exchange features at two points in time have markedly varying effects, from an initial transaction use (time one) to a second transaction occurring two weeks later (time two). Our research objective is to test whether two system design features have the same effects on the intent to continue using an exchange in time two as they had in time one. the two features are control transparency (the availability of information cues) and interim shipping outcome feedback. these effects are mediated, in varying degrees, by perceived information quality. We use social exchange theory and social cognition theory to develop hypotheses regarding changes between time one (the first user transaction) and time two (the second transaction). these are tested using a combined experiment and survey. Supporting the theory, outcome feedback matters 270 NIcOlAOu AND McKNIgHt at time two even though it did not matter at time one. While control transparency has direct effects on a user’s intent to continue use of the exchange in time one, its effects are reduced in time two if negative outcome feedback is communicated to the user. Outcome feedback’s effects grow stronger from time one to time two vis-à-vis control transparency’s effects. this underscores how critical it is to examine such phenomena at more than one period of time. the study also suggests different strategies for managing data exchanges based on the time frame of use. At the initial transaction use, the exchange should make transparent high-quality information cues to its user. At the next transaction, it should provide feedback showing properly fulfilled orders. these findings have implications for both future research examining effective data exchange design and for professionals who wish to enrich electronic data exchange interactions.