ABSTRACT
In general, the choice of the design is based on feasibility, ability to achieve the study goals, and sample size requirements. This chapter reviews design considerations for studies with correlated data and focuses specifically on sample size estimation. It also reviews the importance of careful study design for experiments and observational studies with repeated measures and lists key aspects that need consideration. The chapter describes the most commonly used repeated measures designs and weighs in on their advantages and disadvantages. It introduces methods for sample size calculations for cross-sectional data, and focuses on sample size calculations for clustered and longitudinal data with emphasis on mixed regression and GEE models. The chapter discusses methods based on traditional ANOVA-based approaches and power calculations based on summary measures. It also reviews common randomization approaches for experimental studies, and concludes with a summary and discussion.
