Abstract
Effect sizes are important for power analysis and meta-analysis. This has led to a debate on reporting effect sizes for studies that are not statistically significant. Contrary and supportive evidence has been offered on the basis of Monte Carlo methods. In this article, clarifications are given regarding what should be simulated to determine the possible effects of piecemeal publishing trivial effect sizes.
DOI
10.22237/jmasm/1051748460
Included in
Applied Statistics Commons, Social and Behavioral Sciences Commons, Statistical Theory Commons
