Statistical Analysis of an Adaptive Concept Inventory in Introductory Electric Circuits for Students and Instructors.

Concept assessment instruments utilized in electrical engineering education are primarily designed for students by teaching faculty or assessors of the course. Due to the need for identifying inaccurate understanding early at the introductory level, creating a concept inventory that assesses not just students for their knowledge in fundamental electric circuits but also helps novice teaching faculty identify potential inaccuracies in their assumed correct conceptual understanding. This adaptive mechanism is deemed useful for teaching fundamental concepts effectively. Forty-four undergraduate students and teaching faculty from four academic institutions participated in this study. The collected data were analyzed through descriptive and inferential statistical approaches: item-wise difficulty and discriminatory analysis, inter-item correlations, internal consistency reliability using Kuder-Richardson 20 (KR-20) metric, and exploratory factor analysis. Initial findings from the analysis suggested that the instrument attained acceptable validity, yet its reliability could be further improved. The only significant predictor of the scores was the type of participant (faculty or student). In due course, the outcomes of this study will identify which of the test items significantly contribute to achieving the learning of the instrument's electric circuit concept groups and which of the items need to be improved and supplemented. The set of statistical methods selected for this study offers promising means to enhance concept inventories and make them valid and usable for improving the teaching of fundamentals of electric circuits. [ABSTRACT FROM AUTHOR]