Comparer des méthodes

	Problème des comparaisons multiples	Biais de publication
Domaine	Statistiques de recherche	Statistiques de recherche
Famille	Process / pipeline	Process / pipeline
Année d'origine	1935	1979
Auteur d'origine	Carlo Bonferroni; Benjamini & Hochberg	Robert Rosenthal
Type	Concept	Concept
Source fondatrice	Bonferroni, C. E. (1935). Il calcolo dei coefficienti di correlazione nel caso di variabilità di gruppi. Instituto Italiano di Statistica.	Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86(3), 638–641.
Alias	multiple testing, family-wise error, p-value adjustment, false discovery rate	file drawer problem, selective reporting, outcome reporting bias, funnel plot asymmetry
Apparentées	4	4
Résumé	When conducting multiple statistical tests, the probability of obtaining at least one false positive by chance increases with the number of tests. The multiple comparisons problem (also called the multiplicity problem) occurs because if you conduct 100 hypothesis tests at α = 0.05, you expect ~5 false positives by chance alone, even if all null hypotheses are true. Correction methods—Bonferroni, Benjamini-Hochberg false discovery rate (FDR), and others—adjust the significance threshold or p-values to control error rates. This concept is critical for research integrity and has profound implications for exploratory science.	Publication bias occurs when the results of a study influence whether the study is published. Typically, studies with statistically significant or positive results are more likely to be published than studies with non-significant or negative results, even if both are scientifically valid. This bias distorts the published literature, making treatments appear more effective than they actually are. Rosenthal (1979) termed this the 'file drawer problem': research with null results sits in file drawers, unpublished, creating a biased sample of published evidence. Funnel plots and statistical tests (e.g., Egger test) can detect asymmetry suggesting publication bias; meta-analyses must account for this bias.