Meta-Analysis of Tests for Two Means Using Fixed-Effects Model

This technical report outlines the procedure for calculating power and the number of studies required for fixed-effects meta-analyses involving cluster-randomized designs. It begins by defining meta-analysis and explaining its purpose in systematically assessing multiple studies that yield two group means. The report distinguishes between fixed-effects and random-effects models, emphasizing that the fixed-effects model assumes a common true effect size across studies. The document details the statistical methodologies employed, including power calculations for one-tailed and two-tailed tests, and provides formulas for determining effective sample sizes adjusted for cluster-randomization. An example is included to illustrate the calculation of sample sizes for a specific research scenario, detailing the parameters such as average cluster size and intracluster correlation. The report concludes with numeric results from the calculations, summarizing the outcomes based on the defined hypotheses and parameters.