Sample Size Estimation for Linear Mixed Models with Dependent End Points

Michael Nsiah-Nimo, University of Texas at El Paso

Abstract

The primary objective is sample size estimation in linear mixed model settings. Sample size estimation is an important component of planning a well thought out scientific experiment. Whenever sample size estimation is performed, taking into account a priori model based inferences will provide a sample size estimate that will achieve the desired power without inflating the type I error rate of the study. ^ One common practice is a traditional approach cited in the literature that uses the largest sample size after you Bonferroni the type I error rate to estimate sample sizes as such. We are going to take into account multiplicity using the tree spanning (graph-based) algorithm and improve on just a Bonferroni correction so that we can estimate somewhat lower sample sizes for linear mixed model settings. ^ We present a tree spanning algorithm that is a fast simple novel approach for estimating sample size which focuses on controlling the family-wise error in LMM's with arbitrary dependency structures. This method warrants more powerful bounds compared to the Bonferroni which tends to be more conservative for large set of comparisons. ^ This proposed methodology will yield smaller estimators for sample size may be obtained to make better use of time and resources in experimental settings. ^

Subject Area

Statistics

Recommended Citation

Nsiah-Nimo, Michael, "Sample Size Estimation for Linear Mixed Models with Dependent End Points" (2017). ETD Collection for University of Texas, El Paso. AAI10620057.
https://digitalcommons.utep.edu/dissertations/AAI10620057

Share

COinS