Carl Westin

42 First empirical insights Automation experience External Internal Trust dimensions Performance Process Purpose Automation strategy Dispositional trust Performance shaping factors Workload Skills Strategies Beliefs Control strategy Principles Rules Evaluate ACCEPT REJECT External demands Complexity LOA HMI Context F IGURE 3-1: A functional model of automation usage. external (contextual) factors drive the controller’s binary decision to either use (ac- cept) or disuse (reject) the advice of a hypothetical system. Controller strategies are key to our model, and consist of both control strate- gies on the one hand, 35 and what we term the controller’s automation strategy on the other. This automation strategy can be seen as a trust-related inclination to use the available automation. This strategy is driven by two interactive elements: the overall and general “dispositional trust,” characterized as fairly long-term and sta- ble tendency to either use automation or not, 110, 135 but also a short-term “dynamic trust” that relies on ongoing reappraisal and updated inferences about the system’s predictability and underlying mechanisms. 15 Interaction works in both directions: general tendency can drive beliefs about the underlying mechanisms and motivations of automation, as well as the assump- tions about predicted performance. Moreover, as automation experience accrues, reappraisal informs a recalibration of dispositional trust. That is, more stable, long- term prejudices and biases can be adjusted over the course of interacting with au- tomation. Familiarity with a system’s performance leads to an ongoing reappraisal of the underlying process, for instance. Furthermore as the process model is up- dated, one’s prediction of system performance is similarly reappraised over time.