A Comparison of Logistic Regression Models for Dif Detection in Polytomous Items: The Effect of Small Sample Sizes and Non-Normality of Ability Distributions

Journal Title: International Journal of Assessment Tools in Education - Year 2015, Vol 2, Issue 1

Abstract

This study investigated the effectiveness of logistic regression models to detect uniform and non-uniform DIF in polytomous items across small sample sizes and non-normality of ability distributions. A simulation study was used to compare three logistic regression models, which were the cumulative logits model, the continuation ratio model, and the adjacent categories model. The results revealed that logistic regression was a powerful method to detect DIF in polytomous items, but not useful to distinguish the type of DIF. Continuation ratio model worked best to detect uniform DIF, but the cumulative logits model gave more acceptable type I error results. As sample size increased, type I errors increased at cumulative logits model results. Skewness of ability distributions reduced power of logistic regression to detect non-uniform DIF. Small sample sizes reduced power of logistic regression.

Authors and Affiliations

Yasemin KAYA, Walter L. LEITE, M. David MILLER

Keywords

Related Articles

The Development Study of Thoughts Scale Towards Measurement and Assessment Course on High Education

Considering that measurement and assessment, the most critical component of the education process, should be properly carried out by educators, it is necessary to make sure that education faculties deliver the measuremen...

Assessing Metacognition: Theory and Practices

Many researchers in education emphasized students’ metacognition should be fostered for academic development and achievement. However, to support students’ metacognitive development and adequacy appropriately, their meta...

Evaluating the Comparability of PPT and CBT by Implementing the Compulsory Islamic Culture Course Test in Jordan University

Study aims to determine whether the university students' scores in the compulsory Islamic culture course test on a selected sample differ across the paper-and pencil test (PPT) & computer-based test (CBT) versions, and t...

Exploring Teachers’ Assessment Practices and Skills

The need for increased use of test results to improve educational outcomes is urgent; yet, there is little understanding in the research literature of practitioners’ knowledge and skills in interpreting and using educati...

Determination of Relationship Between the Empathic Tendency Levels and Thinking Styles of Preschool Teacher Candidates

The main objective of this study is to determine the relationship between the empathic tendency levels and thinking styles of preschool teacher candidates. In this study that was patterned with the relational survey mode...

Download PDF file
  • EP ID EP174037
  • DOI -
  • Views 127
  • Downloads 0

How To Cite

Yasemin KAYA, Walter L. LEITE, M. David MILLER (2015). A Comparison of Logistic Regression Models for Dif Detection in Polytomous Items: The Effect of Small Sample Sizes and Non-Normality of Ability Distributions. International Journal of Assessment Tools in Education, 2(1), 22-39. https://europub.co.uk/articles/-A-174037