The Dogmatic Mixture Model Overestimates False Positives

Ulrich Schimmack and Jerry Brunner

No f6y3x, MetaArXiv from Center for Open Science

Abstract: The Bayesian Mixture Model (Gronau et al., 2017) imposes a dogmatic prior on the standard deviation of the z-curve under H1 (for true hypotheses). When actual data have a standard deviation greater than 1, this dogmatic prior leads to inflated estimates of the false positive rate. We demonstrate that false positive estimates decrease when the dogmatic prior is replaced by a free prior. The estimate for cognitive psychology dropped from 40% to 11%. The estimate for social psychology dropped from 62% to 40%. We discuss these results in the context of other meta-analytic models that assume selection for statistically significance.

Date: 2019-04-04
References: Add references at CitEc
Citations:

Downloads: (external link)
https://osf.io/download/5ca66290c1884f0016e982f4/

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:osf:metaar:f6y3x

DOI: 10.31219/osf.io/f6y3x

Access Statistics for this paper

More papers in MetaArXiv from Center for Open Science
Bibliographic data for series maintained by OSF ().