Fail-Safe N is a characteristic value that is used in meta-analyzes . It indicates how many studies with an effect size of zero would have to be found so that the meta-analysis would no longer be significant . Fail-Safe N, however, is not a test statistic , but only a rough guide.
The characteristic value was developed by Rosental (1979) and is based on an idea by Stouffer et al. (1949) back. A calculation method that is not based on the significance but directly on the effect size was described by Orwin (1983).
Purpose of the characteristic
This is to counter the difficulty that many studies are not published if the effect does not become significant and therefore disappears in the scientist's drawer (file drawer problem), whereby the overall significance of an effect is overestimated. This overestimation of the significance is also called publication bias . Literature that is difficult to access is called gray literature .
- meta-analysis . Rainer Hampp Verlag, 2014, ISBN 978-3-86618-975-1 , p. 77 ( limited preview in Google Book search).
- Performance and trust in company partnerships: A meta-analysis on the determining factors for the relationship between performance and trust in company partnerships . kassel university press GmbH, 2011, ISBN 978-3-86219-215-1 , p. 135 ( limited preview in Google Book search).
- Publication Bias in Meta-Analysis: Prevention, Assessment and Adjustments . John Wiley & Sons, 2006, ISBN 978-0-470-87015-0 , pp. 112 ( limited preview in Google Book search).