Correction to “The enhanced research impact of self-archiving platforms: Evidence from bioRxiv”

Abstract

Liu, H., Hu, G., & Li, Y. (2024). The enhanced research impact of self-archiving platforms: Evidence from bioRxiv. Journal of the Association for Information Science and Technology, 75(8), 883–897. https://doi.org/10.1002/asi.24932

This correction has been published following concerns raised by a third party regarding limitations of the data set and analyses presented in the article and the undisclosed modification of the authors’ supporting online dataset following publication.

“The statistical similarity between the treatment and control groups was assessed by a t-test. In our analysis, t-tests were an integral part of the matching procedures, with the null hypothesis being that ‘the variables in the treatment and control groups have equal means’. The results of the t-tests for our treatment and control samples were as follows:

TABLE 1 Results of the t-test for the treatment and control groups.

“The decision to include non-research articles in the control group was based on the fact that bioRxiv deposits include all four types of articles. Since we are examining the overall impact of bioRxiv, it was appropriate to include these article types. Nevertheless, when we excluded non-research articles and reran the regression analysis, the main effect remained significant (0.222***).

“The data acquisition date for Altmetrics differs from that for citation data due to the nature of our study. We measured online exposure between deposition dates and publication dates to estimate its impact. All papers in the treatment group were published before October 2020, and Altmetrics data were collected soon after to avoid potential contamination from web data that could occur over time. In contrast, citation data were collected later, in 2022, with complete citation records up to the end of 2021.”

After being informed about the concerns, the authors made post-publication changes to their original dataset without appropriately stating this in a correction. After acknowledging their mistake, the original, unmodified data set (“Detailed version”) and the consolidated dataset prepared for regression analysis (“Integrated version”) were both made available with an additional explanation regarding the files listed as Supplementary Data. Additionally, the authors have provided the STATA codes used in the analyses, along with article identifiers in the integrated panel dataset to facilitate replication of the study. The authors apologize for this oversight.

Please find the link to the updated supporting dataset, here: https://yunpan.tongji.edu.cn/link/AA5DD75818FCD24946BBCEBEF0C2BAB9CA.