Examining Effect Size Presentations in Introductory Statistics Textbooks for Psychology and Education

• In Event: 68.052 - Weaving Theory in Research Methods

Abstract

Despite policies and recommendations promoting effect size (ES) reporting, recent evidence indicates education and psychology researchers do not consistently follow the practice. Updating previous work, I conducted a content analysis of 28 introductory statistics textbooks published after 2005 to identify characteristics related to the presentation of ES. My findings suggest textbook presentations of ES have become (1) more consistent over time, but they (2) continue to use superficial approaches and (3) rely on small contrived example data sets. These results are interpreted through the lens of a mathematics curriculum framework and in light of Monte Carlo simulations demonstrating an inverse relationship between sample size and effect size variability. The paper concludes with practical recommendations related to textbook selection and creation.

Author

• Andrew Kapral, University of Houston