Diagnostically Cognitive Assessment of Complex Semantic Explanations and Learning: A Bayesian Network Approach

Abstract

This study explored a diagnostically cognitive assessment model for the ANOVA score model emphasizing semantic explanations. The study used the mixed methods designs, in which the ANOVA score model was decomposed into measurable components. This consists of the proficiency student model. Such kinds of data were transferred to a quantitative representation via the Bayesian network model of the ANOVA score model and semantic explanation assessment. This diagnostically cognitive assessment consists of 28 variables hierarchically, which are explanatory variables and evidence variables. Nine variables are explanatory variables that are latent. Nineteen variables are evidence variables that collect students’ learning information and propagate the information to the explanatory variables. The data are simulated data; the semantic explanations from twelve students were recorded and input into the nineteen evidence variables. Semantic explanations indicate 3 levels: lower level, medium level and high level. The score should be more than 82 points, which indicates a mastery level. The study also suggests that if a student achieves a high score in a module, the student has a better chance of achieving a high score in the overall assessment model.