Research reviewers must often assess whether empirical data obtained by a series of psychological studies support the conclusions that an experimental effect is meaningful or not. Recently, several quantitative methods have been developed in order to statistically combine and integrate the evidences across studies. In other words, meta-analysis is an analysis of the results of statistical analysis, and is usually performed on the summary statistics such as sample means, variances, and correlation coefficients reported by many previous studies. In this paper, a Bayesian approach is presented in order to statistically review a series of psychological studies and draw a conclusion based on effect sizes against the conflicting psychological hypotheses. (A)
Abstract