Abstract. This study empirically tests a social psychology-based Information Processing Model (IPM) that explains how events may change trust over time based on three cognitive mechanisms or "gears": attention, attribution, and judgment. We briefly describe the IPM, and then empirically validate its assumptions and extensions. The IPM is contrasted with the incremental growth model (IGM) of trust change. We find more support for the IPM than the IGM.Keywords: Trust, attention, attribution, threshold, risk, illusion, change.
IntroductionHow does trust in another party change in response to a series of events over time? How do cognitive mechanisms produce trust changes over time? Few studies have addressed these questions (e.g., [3]). These questions are vital because trust in a technology changes over time. Further, the world is becoming scarier in terms of how adroitly online agents are attacking targets once considered safe, which changes trust. McKnight et al.[6] addressed these questions by developing a psychology-based information processing model (IPM) of the major cognitive mechanisms involved in trust changes. Then they simulated the model to illustrate how it works under a set of plausible yet untested assumptions. But their assumptions may or may not be right. Incorrect assumptions would render the IPM useless. We test their assumptions.The IPM can serve as a basic way to depict trust changes in the cyber world. By trust we mean a willingness to be vulnerable to another party [4]. See McKnight et al.[6] for the general developmental process theory approach behind this paper and the rationale for the theoretical model being tested. We next briefly describe the IPM. Then we explain the empirical methods used. We report the results, which validate much of the IPM and suggest refinements. This study enhances the model's value.
2The Baseline Information Processing Model
Model OverviewModel action initiates when an event takes place that may have trust ramifications (Figure 1, 1. Event). The event may be an interaction with the trustee, or it may be some signal or second-hand report about the trustee (e.g., media reports).