Could It Be Better to Discard 90% of the Data? A Statistical Paradox

Conventional practice is to draw inferences from all available data and research results, even though there is ample evidence to suggest that empirical literatures suffer from publication selection bias. When a scientific literature is plagued by such bias, a simple discarding of the vast majority of empirical results can actually improve statistical inference and estimation. Simulations demonstrate that, if the majority of researchers, reviewers, and editors use statistical significance as a criterion for reporting or publishing an estimate, discarding 90% of the published findings greatly reduces publication selection bias and is often more efficient than conventional summary statistics. Improving statistical estimation and inference through removing so much data goes against statistical theory and practice; hence, it is paradoxical. We investigate a very simple method to reduce the effects of publication bias and to improve the efficiency of summary estimates of accumulated empirical research results that averages the most precise ten percent of the reported estimates (i.e., ‘Top10’). In the process, the critical importance of precision (the inverse of an estimate’s standard error) as a measure of a study’s quality is brought to light. Reviewers and journal editors should use precision as one objective measure of a study’s quality.