Answers are grounded in this paper and cite their sources.
Meta-analysis explores a new method for converting standardized beta weights to correlations, suggesting improved practices for cumulative sciences.
, correctly identify error variance. Due to imputation error, reporting beta weights alone typically destroys 95% to over 99% of the information originally held by a full correlation matrix. For fields that almost exclusively report betas, such as economics, it necessarily hampers them from becoming cumulative sciences. We recommend that all previous uses of Peterson and Brown be reevaluated, future aggregations of standardized beta weights use our provided imputation techniques, and correlation matrices be routinely reported. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
No takes yet. Share an insight, caveat, or question.
Steel et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: