A meta-analytical answer to the crisis of confidence in psychology

Abstract

Meta-analysis is a firmly established methodology and an integral part of the process of generating knowledge across the empirical sciences. Meta-analysis has also focused on methodology and has become a dominant critic of methodological shortcomings. We highlight several problematic issues on how we research in psychology: excess of heterogeneity in the results and difficulties for replication, publication bias, suboptimal methodological quality, and questionable practices of the researchers. These and other problems led to a “crisis of confidence” in psychology. We discuss how the meta-analytical perspective and its procedures can help to overcome the crisis. A more cooperative perspective, instead of a competitive one, can shift to consider replication as a more valuable contribution. Knowledge cannot be based in isolated studies. Given the nature of the object of study of psychology the natural unit to generate knowledge must be the estimated distribution of the effect sizes, not the dichotomous decision on statistical significance in specific studies. Some suggestions are offered on how to redirect researchers' research and practices, so that their personal interests and those of science as such are better aligned.