Despite decades of warnings, many areas of science still insist on labelling a result of P < 0.05 as “significant”. In this article, D. Colquhoun discusses whether the inconsiderate use of p-values and other questionable research practices like multiple comparisons, lack of randomisation and P-hacking account for a substantial part of the reproducibility issue. He concludes that science is endangered by statistical misunderstanding, and by people who impose perverse incentives on scientists.
1 Comment
Great article! Thank you for addressing the issue of misinterpretation of P values and its impact on research reproducibility. It's crucial to highlight the pervasive nature of questionable research practices and the need for greater statistical understanding. Your insights shed light on the dangers faced by science due to these practices and the imposition of perverse incentives. Keep up the excellent work!
Leave A Comment