The file drawer problem and tolerance for null results

By R. Rosenthal. 1979

Subfields: Statistics and Probability, Social Sciences & Humanities

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

In Social Sciences & Humanities

Citations in the corpus

When is multitask learning effective? Semantic sequence prediction under varying data conditions (2016, background): “In fact, results on semantic tasks are either mixed (Collobert et al., 2011) or, due to the ﬁle drawer bias (Rosenthal, 1979), simply not reported.”
Automated versus Do-It-Yourself Methods for Causal Inference: Lessons Learned from a Data Analysis Competition (2017, background): “Understanding of the relative performance of methods can be biased due to the “file drawer effect” (Rosenthal, 1979).”
Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. (2006, result): “On the one hand, studies with large effect sizes may have a better chance of publication compared with studies with small or nonsignificant effects (the " file drawer problem " ; Rosenthal, 1979).”
On the use of beta coefficients in meta-analysis. (2005, background): “This is commonly referred to as the “file drawer problem” (Orwin, 1983; Rosenthal, 1979) and is typically addressed by estimating the number of studies with an effect size of zero that would be necessary to reduce a particular relationship to nonsignificance.”
Impact of informal institutions on the prevalence, strategy, and performance of family firms: A meta-analysis (2020, methodology): “To control for methodological artifacts, we tested for the ‘‘file drawer problem’’ (Rosenthal, 1979) by including a dummy variable indicating whether a study was published or not (reference group).”
A meta-analysis of bees' responses to anthropogenic disturbance. (2009, methodology): “Finally, we used Rosenberg’s 2005 fail-safe number calculator to estimate the number of nonsignificant, unpublished studies that would need to be added to a meta-analysis to nullify its overall effect size (Rosenthal 1979).”
Research Design in Qualitative/Quantitative/Mixed Methods (2011, background): “Quantitative methods have a long history, dating to at least the 1930s, that has produced strong professional norms that impact research activities, such as the criteria used to make funding decisions (What Works Clearinghouse, 2008) and decisions about the kinds of studies and results likely to be ”
Synthesizing and Extending Resource Dependence Theory (2013, unknown): “…2013 measurements (reference category); (4) publication status: “published” or “unpublished” (reference category), to test for the presence of file drawer problems (Rosenthal, 1979); (5) publication year: to correct for otherwise unmodeled time effects; (6) performance type: “legitimacy,”…”
Does Burnout Affect Academic Achievement? A Meta-Analysis of over 100,000 Students (2020, methodology): “To do so, the fail-safe N statistic was calculated to estimate the number of unpublished studies with null findings that would be necessary to reduce the effect size to zero (Rosenthal, 1979).”
A meta-analytic review of research on gender differences in sexuality, 1993-2007. (2010, background): “In many meta-analyses, publication bias and the file drawer effect are concerns (Rosenthal, 1979).”