All articles tagged with #metascience
A study analyzing 18,400 articles suggests that peer reviewers are more likely to approve manuscripts if their own work is cited, raising concerns about citation bias and coercive practices in peer review. Reviewers who request citations of their own work tend to be less likely to approve the article, and language used in reviewer comments may indicate coercion, highlighting potential issues in the peer review process.
A study of over 18,000 effect sizes from more than 350 published studies in ecology reveals evidence of selective reporting and exaggeration of effect sizes. The research highlights the prevalence of underpowered studies and biased reporting in the field of ecology, consistent with previous meta-science studies.
Fraud in science is becoming increasingly common, and automated screening tools are not enough to tackle it. While they have been successful in detecting faulty experiments and plagiarism, they can be easily fooled by fraudsters. The solution is to employ more people to check the quality of research, such as the Dutch microbiologist Elisabeth Bik, who has caught multiple fraudsters. Quality control systems need not be expensive, and random spot checks could be effective. Institutions and funders should be notified of any wrongdoing, and there should be consequences for repeat offenders.