Application of Logistic Regression Model in Physics Education

Abstract

This paper introduces a logistic regression model used to analyze multiple-choice test data of physics test. It does not involve decision making. It is a predictive model and is more akin to nonlinearity, such as fitting a polynomial to a set of data values. Brief description of the goals and algorithms of such a model is provided, together with examples illustrating their applications in physics. Priority haven been given for data interpretation rather than mathematical complexities.

Key words: Logistic regression; Multiple choice test items; Classical test theory; Item response theory; Probability; Evaluation

The Himalayan Physics

Vol.1, No.1, May, 2010

Page: 4-9

Uploaded Date: 28 July, 2011