In a simulation study, we generated repeated measures data with various parameters (e.g., sample size, between- and within-individual level variance, true effect size, predictor distribution, etc.). We applied multiple methods to the simulated data, to compare their performance with regards to power, false positive rate, and estimate accuracy.
The methods investigated were naive linear regression, aggregate regression, cluster robust standard errors, fixed effects models, linear mixed models, and generalized estimating equations
The complete results of the simulations are available for viewing and download at the following Google Drive link.
For more information, please contact by [email protected] or through GitHub Platform.