学术文献库

Brewe, Bruun, and Bearden first applied network analysis to understand patterns of incorrect conceptual physics reasoning in multiple-choice instruments introducing the Module Analysis for Multiple-Choice Responses (MAMCR) algorithm. Wells {\it et al.} proposed an extension to the algorithm which allowed the analysis of large datasets called Modified Module Analysis (MMA). This method analyzed the network structure of the correlation matrix of the responses to a multiple-choice instrument. Both MAMCR and MMA could only be applied to networks of incorrect responses. In this study, an extension of MMA is explored which allows the analysis of networks involving both correct and incorrect responses. The extension analyzes the network structure of the partial correlation matrix instead of the correlation matrix. The new algorithm, called MMA-P, was applied to the FCI and recovered much of the structure identified by MMA. The algorithm also identified sets of correct answers requiring similar physical reasoning reported in previous studies. Beyond groups of all correct and all incorrect responses, some groups of responses which mixed correct and incorrect responses were also identified. Some of these mixed response groups were produced when a correct response was selected for incorrect reasons; some of the groups helped to explain the gender unfairness previously reported for the FCI.