Biostatistics and Bioinformatics Monthly Seminar-October

“Repeated Measures Designs in Cancer Studies”

Li Liu, PhD
Biostatistician, UI Cancer Center
Associate Professor, Division of Epidemiology and Biostatistics
School of Public Health, University of Illinois at Chicago

Register via Zoom here

ABSTRACT:

In repeated measures design, when a covariate was repeatedly measured, it is typically treated the same way as any covariates in mixed-effects regression models. In this presentation, I will demonstrate that a repeatedly measured covariate can have effects at both the individual level and the repeated measure level. The assumption that these two levels of effects are equal was typically made unknowingly. Failure to check this assumption can lead to dubious conclusions about the effect of the repeatedly measured covariate.

I will also demonstrate the efficiency of repeated measures design using an animal cancer research data. In particular, the power of such design depends on the number of repeated measures, as well as the size of correlations between repeated measures.

Share on facebook
Facebook
Share on twitter
Twitter