An investigation into the nature and constructs of relative expertise in economic problem-solving.

摘要

Relying upon the research literature on expert problem solving, this study investigated the nature and constructs of relative expertise in economic problem solving. The study examined literal transcripts generated from 'talk-aloud' protocols of twenty-eight participants responding to three economic problems. Each of the following sub-groups contained four participants: high school students who had not taken economics, high school students who had, undergraduate non-majors, undergraduate economic majors, graduate students in economics and two groups of Ph.D. economists. Protocols were coded using thirteen indicators of expertise in economic problem solving as identified in the research literature and by previous study examining expert problem solving in economics (Miller and VanFossen, 1994). Mean levels of each of the thirteen indicators were analyzed using one-way analysis of variance techniques to determine if differences across sub-groups could be identified. Factor analysis was conducted to determine if several underlying variables might be identified. The data were re-coded using a standardized ranking technique and these results were compared across sub-groups. Cluster analysis was conducted to determine if similarities existed among particular sub-groups. The study employed an ordered Probit analysis to develop a model for scaling transcribed responses to economic problems, the Expertise in Economic Problem Solving Scaling Model (EEPSSM). Results of these analyses indicated that it was possible to identify a group of economic problem solvers between the extreme categories of expert and novice: relative experts in economic problem solving. Factor analysis indicated two underlying variables: Economic Knowledge and Economic Problem Solving Processes. The EEPSSM scaled the sample at least as well as the standardized rankings. The analysis also indicated a very strong, positive relationship between each of the thirteen indicators of expertise and participant's category of expected expertise. Results suggested that, for this sample, pre-collegiate, and to some extent undergraduate, economic education had little impact on demonstrated level of expertise in economic problem solving. Finally, the data indicated overwhelmingly that the acquisition of expertise in economic problem solving was not a linear continuum. In fact, participants demonstrated a qualitative 'leap' in expertise at points along the path to expertise.