Choosing a Statistical Model Amidst Ceiling and Floor Effects

This session is from the NASP 2021 Annual Convention

  • PMD: 1,9
  • Skill Level: Introduction
  • Session 6: Recent Developments in Assessment and Research

Learner Objectives

This session will help participants…

  1. identify ceiling and floor effects in their data.
  2. determine which statistical models to use when estimating group differences.
  3. learn how to model group differences in the presence of ceiling and floor effects.


School psychology research often entails group comparisons, commonly modeled by analysis of variance (ANOVA) based models. However, these models underestimate true differences between groups when ceiling or floor effects (CFE) are present. To guide researchers, we conducted a simulation study to analyze bias in ANOVA and censored regression models when estimating mean differences between groups. Researchers will learn when and how to use each method through an illustrated example based on a class-wide academic intervention.



Vimal V. Rao, University of Minnesota
Kristin E. Running, University of Minnesota
Robin S. Codding, Northeastern University

Letter to Supervisor

Looking for a concise way to pitch the convention to your employer? This letter and the corresponding talking points are here to help!

Preliminary Brochure

Share our NASP 2022 preliminary brochure with your supervisor along with the above letter.