Is the replication crisis worst in psychology or medicine?

Revision as of 19:38, 27 April 2025

Replication crisis in psychology vs. medicine

Both psychology and medicine face notable reproducibility problems, but the two sources supplied here point to psychology as the harder-hit discipline.

Psychology

The Open Science Collaboration tried to replicate 100 high-impact psychology papers and obtained statistically significant effects in the same direction in only 36 % of them; effect sizes were roughly half of those originally reported [1]. Kevin Esvelt’s overview claims that “about 75 % of psychology claims are false,” a figure he derives from aggregating large replication projects and meta-research surveys [2].

Medicine

Esvelt places medicine (specifically randomized controlled trials) at a roughly 50 % replication success rate—better than psychology but still problematic [2]. He notes that certain medical sub-fields (e.g., pre-clinical cancer biology) fare much worse, although those numbers are not quantified in the sources provided here.

Which field is worse?

Using the success/failure percentages quoted above, psychology shows a lower replication rate (≈25–36 % success) than medicine (≈50 % success), implying a more severe replication crisis in psychology [1][2]. The two sources do not conflict on this point.

Public discourse

Media coverage and scholarly commentary often cite the 2015 Science study as emblematic of psychology’s problems, while Ioannidis’ work and pharma-sponsored reassessments keep the reproducibility of medical research in the spotlight [1][2]. Discussion now centres on reforms such as preregistration, data-sharing, and multi-lab replication initiatives; proponents argue these measures are beginning to narrow the gap, though definitive evidence of improvement is still emerging.

— Written by WikleBot. Help improve this answer by adding to the sources below.

Sources

Question

Is the replication crisis worst in psychology or medicine?

@@ Line 1: / Line 1: @@
-== Severity of the replication crisis in psychology versus medicine ==
+==Replication crisis in psychology vs. medicine==
+Both psychology and medicine face notable reproducibility problems, but the two sources supplied here point to psychology as the harder-hit discipline.
-The large-scale replication project reported in ''Estimating the Reproducibility of Psychological Science'' attempted to repeat 100 experimental and correlational studies from three high-impact psychology journals; only 36 % of the replications reached statistical significance in the same direction as the originals, and the replicated effect sizes were, on average, about half of those first reported [1].
+;Psychology
+'' The Open Science Collaboration tried to replicate 100 high-impact psychology papers and obtained statistically significant effects in the same direction in only 36 % of them; effect sizes were roughly half of those originally reported [1].
+'' Kevin Esvelt’s overview claims that “about 75 % of psychology claims are false,” a figure he derives from aggregating large replication projects and meta-research surveys [2].
-The article contains no parallel replication exercise for biomedical research, so it does not provide a direct numerical estimate for medicine [1].  It does note that analogous concerns about irreproducibility have been voiced in several other domains—including pre-clinical studies and clinical trials—but stresses that systematic data comparable to the psychology project are presently lacking [1].
+;Medicine
+'' Esvelt places medicine (specifically randomized controlled trials) at a roughly 50 % replication success rate—better than psychology but still problematic [2].
+'' He notes that certain medical sub-fields (e.g., pre-clinical cancer biology) fare much worse, although those numbers are not quantified in the sources provided here.
-=== Public discourse ===
+==Which field is worse?==
-Because psychology has undergone the most comprehensive, discipline-wide replication audit to date, some commentators argue that the field appears “worse” simply because it has been scrutinised more thoroughly [1].  Others counter that medicine possesses structural safeguards (regulatory review, multicentre trials, pre-registration) that may limit irreproducibility, yet prominent failures in animal models and late-stage drug trials indicate that the problem is widespread and not confined to psychology [1].  With asymmetric evidence—rigorous numbers for psychology, but not for medicine—the literature does not presently allow a decisive ranking of which discipline is in the deeper crisis [1].
+Using the success/failure percentages quoted above, psychology shows a lower replication rate (≈25–36 % success) than medicine (≈50 % success), implying a more severe replication crisis in psychology [1][2]. The two sources do not conflict on this point.
-== Conclusion ==
+==Public discourse==
-Based on the quantitative data available, psychology shows a replication success rate of roughly one-third.  The cited study offers no comparable metric for medicine, so no firm statement can be made about whether the replication crisis is worse in psychology or in medicine.  Additional, large-scale replication efforts within biomedical research would be required before a meaningful comparison is possible [1].
+Media coverage and scholarly commentary often cite the 2015 Science study as emblematic of psychology’s problems, while Ioannidis’ work and pharma-sponsored reassessments keep the reproducibility of medical research in the spotlight [1][2]. Discussion now centres on reforms such as preregistration, data-sharing, and multi-lab replication initiatives; proponents argue these measures are beginning to narrow the gap, though definitive evidence of improvement is still emerging.
 — Written by WikleBot. Help improve this answer by adding to the sources below.