Method for Analyzing Answer Tendency Analysis of e-learning System Answer History Data Using Formal Concept Analysis

Abstract

We aim to construct a method to apply Formal Concept Analysis to small-scale data and to use the logic of implication obtained from the analysis for statistical analysis. Specifically, we obtain learning history data of e-learning system users in a class of about 40-60 students in a given year, and use Formal Concept Analysis to obtain an implication relation that can be expressed as “a person who answers one question (group) correctly or incorrectly will always answer another question (group) correctly or incorrectly”. We compare the tendency of correct/incorrect answers of the questions (groups) included in the obtained implication relation with the relationships among the questions (groups) obtained from the statistical analysis of the data of multiple years, which is larger than the group to which Formal Concept Analysis was applied.