Cognitive Task Analysis as an Empirical Framework for Identifying Cross-Domain Training Opportunities

• In Event: 44.038 - Learning From Each Other: Can We Better Achieve Social Goals With Cross-Professional Teams?

Abstract

Finding the knowledge overlap amongst different fields and professional endeavors may be beneficial for the purpose of realizing efficiencies in training and the dissemination of relevant knowledge to practitioners. Currently, practitioners tend to value only research on training that is conducted within their specific domain and may not recognize where research on training in a different field could in fact be valuable. Such lay decisions are to be expected, because expertise (the goal of professional training) is, by definition, specific to a domain (Chi et al., 1988; Ericsson & Smith, 1991). Thus, domain identification sharpens focus on the distinct contexts in which expert performance is both executed and evaluated. However, the identification of discrete domains has been an under-explored facet of the expertise research. While there is extensive and consistent evidence that the skilled performance constituting expertise is reliably demonstrated only within the particular domain in which the skills have been developed, there has been very little examination of the defining properties of domains themselves that could meaningfully guide cross-domain training.

By default, domains have been defined through the observational grouping of knowledge and performances associated with a socially-defined activity (Wertsch, 1991). School subjects and occupations, for example, are customarily considered different domains on this basis (Barnett & Ceci, 2002). However, these inexact differentiations raise problematic issues for delineating training. For example, Wineberg (1998) compared the performances of two expert historians with different specializations in American history on a task related to Abraham Lincoln. The Lincoln specialist utilized more refined strategies in his analysis and demonstrated stronger overall performance, indicating that the domain of “American history” was not a sufficient domain boundary definition. Further, Barnett and Koslowski (2002) compared the performance of expert restaurateurs and business consultants without restaurant-specific knowledge or experience on a restaurant business simulation that required the use of problem-solving skills “at the intersections of the [restaurant and business] domains” (p. 237). The business consultants exhibited better performance, despite the assumption that their expertise lay outside the restaurant domain. These examples suggest that there was poor alignment between the subjects’ actual and postulated domains of expertise.

Cognitive task analysis (CTA) provides an empirical underpinning to identify these overlaps and differentiations based on structural similarities across or differentiations within domains. CTA is a collection of knowledge elicitation and analysis strategies for capturing expert knowledge, including the tacit knowledge that differentiates experts from non-experts (Feldon, 2007). Beyond simply eliciting knowledge, the practice of CTA entails analysis and representation of what is elicited (Crandall et al., 2006)—typically resulting in the identification of discrete concepts, skills, and decision points with accompanying cues that comprise domain expertise (Clark et al., 2008). Such classifications permit empirical analyses of domain structures (e.g., process- vs. product-oriented; Ceci, 1989) and competencies to identify areas of overlap. Doing so would provide an evidence-based strategy for grouping professional competencies across domains for the purposes of realizing efficiencies in the presentation of relevant information to stakeholders with diverse professional identities whose training strategy and content needs are similar.

Author

• David F. Feldon, Utah State University