Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 155 - Contributed Poster Presentations: Mental Health Statistics Section
Type: Contributed
Date/Time: Monday, August 8, 2022 : 10:30 AM to 12:20 PM
Sponsor: Mental Health Statistics Section
Abstract #323514
Title: Accurate Confidence Interval Estimation for Non-Centrality Parameters and Effect Size Indices
Author(s): Kaidi Kang* and Kristan Armstrong and Suzanne Avery and Maureen McHugo and Stephan Heckers and Simon N Vandekar
Companies: Vanderbilt University and Vanderbilt University Medical Center and Vanderbilt University Medical Center and Vanderbilt University Medical Center and Vanderbilt University Medical Center and Vanderbilt University Medical Center
Keywords: effect sizes; robust effect size index; confidence interval; non-centrality parameters; bootstraps

Effect size indices are useful tools for communicating study findings. Reporting effect size index estimates with their confidence intervals (CIs) can be an excellent way to simultaneously communicate the strength and precision of the observed evidence. We recently proposed a robust effect size index (RESI) that is advantageous over common indices because it’s widely applicable to different types of data. In this research, we use statistical theory and simulations to develop and evaluate RESI estimators and CIs that rely on different covariance estimators. Our results show (1) counter to intuition, the randomness of covariates reduces coverage for Chi-square and F CIs; (2) when the variance of the estimators is estimated, the non-central Chi-square and F CIs using the parametric and robust RESI estimators fail to cover the true effect size at the nominal level. Using the robust estimator along with the proposed bootstrap CI provides valid inference for the RESI, especially when model assumptions may be violated. We propose a framework for the analysis of effect size (ANOES), such that effect sizes with CIs can be easily reported in an analysis of variance (ANOVA) table format.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2022 program