Carl Westin

28 Strategic Conformance: A Literature Survey 2-3-3 Synthesizing acceptance models across communities Two separate research groups recently presented theoretical acceptance models syn- thesizing the TAM and IDT frameworks with relevant knowledge on automation use derived from CE research to create the Automation Acceptance Model (AAM) 64 and the Adjusted Automation Acceptance Model. 14 While similar in the sense that both models rely on TAM architecture, they diverge in the additional and modi- fied elements considered. Both models highlight a concern within the CE domain that current models, and individual drivers of acceptance, do not adequately capture acceptance. The AAM model specifically embraces the compatibility construct, along with trust, as key drivers for automation acceptance (Figure 2-1). The feedback lines denote that acceptance is a bi-directional process. Consider the relationship between trust and acceptance. In order to start using a new system, there has to be trust. Yet trust, specifically for that system (to be distinguished from dispositional trust 110 ), cannot be built without using the system. Compatibility on the other hand, is a perceived characteristic of the automation that can influence trust. While the CE community has predominantly contributed with task-technology compatibility in cognitive systems and the dynamics of trust and reliance, 64 ac- ceptance theories originated in the information systems community have provided coherent acceptance frameworks, including clear methodologies for measuring ac- ceptance. The divergence in research focus between the two communities has been attributed to differences in research granularity. 64 While the CE community has mainly focused on short-term, micro-level observations of operator behavior, the information systems community has approached acceptance by looking at human- F IGURE 2-1: The Automation Acceptance Model (adapted from 64 ).