Is kinship related to retractions? A case-control study of retracted publications with repeated surnames

The aim of this study is to investigate the relationship between co-authorship with family members and article retractions. To this end, an experimental model was designed comparing the proportion of articles with co-authors sharing the same surname in a control sample and an experimental sample. The control sample was randomly selected using OpenAlex, while the experimental sample was obtained from RetractBase, a database specializing in retracted publications. The probability that two authors share a surname by chance was calculated by considering the frequency of the surname in the scholarly literature and the number of authors per article. The results show that, at the most restrictive probability threshold (<0.01%), no statistically significant differences were observed. However, among publications from Western regions (Europe, North America, and Oceania), there is a higher proportion of retracted publications with repeated surnames than in the control group, particularly in English-speaking countries. Publications in disciplines associated with biomedicine show a higher proportion of repeated surnames than those in the control group, while the proportion of retracted publications with repeated surnames appears to decrease over time. The most common reasons for retraction are authorship issues and ethical issues. The results suggest that the larger differences observed in Western countries could reflect the lower prevalence of these practices in these regions, combined with a high retraction rate among publications involving potentially illegitimate authorship.