Do retraction practices work effectively? Evidence from citations of psychological retracted articles

Abstract

Scientific retraction practices are intended to help purge the continued use of flawed research and assist in maintaining the integrity, credibility and quality of scientific literature. However, the practical effect of retraction is still vague and needs to be further explored. In this study, we analysed the citation counts and sentiments (positive/negative) of retracted articles in psychology journals from Web of Science to explore the effect of retraction. Causal inference strategies were used to measure the net effect of retractions on citation. Results show that the retraction practices induced the citation counts to reduce as expected. However, the proportion of negative citations also decreased because of retraction, indicating an unsatisfied effect. The retraction practice of high-impact factors and open access journals was more effective than other journals. The study integrated an understanding of the dissemination of erroneous publications and provided implications for liabilities involved in the whole retraction process.

Self-correction in science: The effect of retraction on the frequency of citations

“…However, we also found a strong and direct loss of citations due to retraction, cutting citations down to less than half during the following years. These findings are in line with others, also reporting retraction losses [ 4 , 22 , 24 , 28 ], and the size of the loss is comparable over studies. Distinguishing between papers that are retracted shortly after publication and papers that are retracted more than two years after publication, shows that the retraction penalty exists in both phases of a scientific paper.…”

“…Note also that some of the citations to retracted papers are negative, citing the paper because it was retracted rather than as valid evidence. Limited evidence indicates that the frequency of negative citations seems to be rather low, being well below 10% [ 7 , 10 , 24 ].…”

“…These studies have small samples or are outdated now. Thus, sceptics remain skeptical, mostly because of the recent findings of continued affirmative citation after retraction [ 7 , 24 , 25 ]. Decisive evidence is necessary.…”

“…To evaluate the effect of retractions, a control group of non-retracted papers undergoing the same citation pattern is needed. Some authors [e.g., 9,17,[22][23][24] did such studies. Their control sample consisted of papers published immediately before and after the retracted paper.…”

Characteristics of correction practice and its citation in library and information science journals

“…The effectiveness of correction practice and retraction practice is an important topic that can help us understand the importance of these two mechanisms. For the retraction mechanism, several studies showed that retracted articles continued to be cited after their retraction (da Silva and Bornemann-Cimenti, 2017; Yang et al, 2022), and the analysis of the nature of the citations revealed that most of them were positive (Bar-Ilan and Halevi, 2017; Bolboacă et al, 2019; Suelzer et al, 2019). In addition, Shamsi et al also explored the extent to which retracted articles were shared via social and mass media.…”