Individual Swift Trust and Knowledge-Based Trust in Face-to-Face and Virtual Team Members

traditionally, trust has been seen as a result of personal knowledge of an individual’s past behavior. In this view, trust develops gradually over time based on an individual’s cognitive assessment of the other person’s behavior. However, high levels 242 RObERt, DENNIS, AND HuNg of trust have been observed among members of virtual teams, who often have little prior history of working together and may never meet each other in person. to integrate these two seemingly contradictory views of trust, this study manipulated team member characteristics and team member behavior to empirically test a two-stage theoretical model of trust formation and the influence of information and communication technologies (ICt) on trust formation. the results indicate that category-based processing of team member characteristics and an individual’s own disposition to trust dominated the initial formation of swift trust. Once individuals accumulated sufficient information to assess a team member’s trustworthiness, the effects of swift trust declined and knowledge-based trust formed using team members’ behaviors (perceived ability, integrity, and benevolence) became dominant. the use of ICt increased perceived risk of team failure, which reduced the likelihood that team members would engage in future trusting behaviors. KeY woRDs AnD PHRAses: cognitive trust, computer-mediated communication, initial trust, knowledge-based trust, presumptive trust, swift trust, trust, vignettes, virtual team. tRust is An essentiAL eLement in both organizational and team phenomena. trust affects performance [53, 59, 71], and even more so when tasks are highly interdependent [2, 67]. trust is a critical element in enhancing intraand interorganizational cooperation, coordination, and control [64, 77, 94]. trust allows individuals to justify their decision to contribute [15, 76, 78] and allows individuals to freely exchange information and share knowledge that is critical to the success of collaboration [35, 60]. trust is even more important in virtual teams, in which members communicate mainly through information and communication technologies (ICts) and are often geographically dispersed. traditional control mechanisms imported from a face-to-face communication environment have been shown to be less effective in an ICt-mediated communication environment [114]; thus, control based on authority is often yielded to self-direction and self-control [53], which relies heavily on trust. trust is an individual-level construct and is often defined as an individual’s willingness to be vulnerable to the actions of another person [94]. trust has traditionally been seen as the result of individual judgments of past behavior in which derived costs and benefits of such future behavior are cognitively assessed [75, 149]. based on positive outcomes of repeated behavior, individuals develop trust gradually over time [6, 82, 94]. this traditional view of trust would predict low levels of initial trust in virtual teams because team members have little past history, may not share common cultures, and use ICt extensively, which can limit direct personal observations that allow members to perform effective cognitive trust assessment [21, 116]. However, high levels of initial trust called swift trust have been observed among members of temporary teams [99] and virtual teams [61, 63]. We believe that these seemingly contradictory views of trust call for a reexamination of trust formation. First, we argue that these two views of trust—knowledge-based trust developed through interactions and swift trust developed prior to interaction—are INDIVIDuAL SWIFt tRuSt AND KNOWLEDgE-bASED tRuSt 243 two different forms of trust that are formed through fundamentally different processes. unlike previous empirical studies of trust in virtual teams [61, 63, 67], we follow McKnight et al. [98] and Meyerson et al. [99] and treat swift trust as a categorymatching process based on team member characteristics, not on their behaviors, and thus we measure swift trust prior to any knowledge of team members’ behavior. Knowledge-based trust is based on the assessment of behavior and thus we measure it after subjects have been exposed to past behaviors of future team members. Second, we agree that initial swift trust perceptions are fragile [83, 98, 99], but we argue that these initial, fragile—and often wildly inaccurate—perceptions have a greater impact on subsequent trust than previously believed. these initial judgments influence how individuals perceive the outcomes of a trusting behavior and how they interpret behaviors that either support or refute that judgment [10, 75, 153]. Finally, we argue that the communication environment through which team members interact (face-to-face or ICt mediated) influences a team member’s willingness to engage in trusting behavior. ICt environments slow the relationship development processes [2, 18, 26, 143, 145] and lack traditional monitoring mechanisms (e.g., direct supervision), which reduces the sense of social control and coordination [63, 114]. both of these factors increase the perception of risk associated with team outcomes. Prior theories and the Proposed Model