International Congress on
Peer Review and Scientific Publication

Hong Chen,¹ Christopher I. Rider,² David Jurgens,^1,3 Misha Teplitskiy¹

Objective

This study examines which intended scientific contributions become published contributions. While many manuscripts are rejected, there is limited understanding of which researchers successfully navigate rejection and resubmission. We investigate geographical disparities in postrejection publishing outcomes, focusing on differences between authors from Western and non-Western countries. We investigate several potential mechanisms underlying these disparities, especially procedural knowledge of how to interpret editorial decisions for rejection, revision, and resubmission.

Design

Partnering with the Institute of Physics Publishing, we collected metadata on 126,000 rejected submissions made in 2018 through 2022 from 62 peer-reviewed journals. We tracked whether, when, and where these rejected manuscripts were eventually published by developing a classifier that we validated with ground-truth data provided by the authors of rejected manuscripts. We defined Western authors as those affiliated with (self-reported) institutions in North America, Europe, or Oceania, and non-Western authors as those from Asia, Africa, and Latin America. The analysis examined the publishing disparities based on author geography while controlling for key confounders—perceived submission quality, proxied by peer reviewer evaluations at the time of rejection (including desk rejection)—as well as time, team composition, and corresponding author characteristics, such as gender, seniority, and current institutional affiliation ranking. The prior Western coauthors and Western affiliations were also included, serving as proxies for access to procedural knowledge. To further probe procedural knowledge as a mechanism, we examined multiple dimensions of how authors interpret and respond to rejection in a survey of corresponding authors whose manuscripts were rejected.

Results

Among the authors included, 43% had Western affiliations and 57% had non-Western affiliations. Authors from Western countries had better postrejection outcomes. Despite similar perceived quality at the time of rejection, Western authors were 6.7% more likely to publish the rejected papers elsewhere. Among those who ultimately published (59.3% of all rejected papers), Western authors published approximately 23 days faster and did 5.9% less abstract revising and 12.0% less authorship team change. We also found evidence suggesting that greater access to procedural knowledge, proxied by prior publishing experience and prior Western affiliation and coauthorship, contributed to better postrejection outcomes. These factors explain a substantial portion of the observed disparities in postrejection publication rates, more so than perceived submission quality or institutional resourcefulness. We also surveyed 10,000 corresponding authors of rejected manuscripts in June 2023 and received 287 responses. Limited by the very low response rate, survey responses indicated no significant differences between Western and non-Western authors in their perceptions of feedback fairness or negativity, planned revisions, or target outlet prestige, suggesting that other dimensions of procedural knowledge might be key.

Conclusions

This study identified geographical disparities in the file drawer of science. Western authors experienced better postrejection outcomes than non-Western authors, and the findings suggest that differential access to procedural knowledge may contribute to these disparities.

¹School of Information, University of Michigan, Ann Arbor, MI, US, tepl@umich.edu; ²Ross School of Business, University of Michigan, Ann Arbor, MI, US; ³Department of Computer Science and Engineering, University of Michigan, Ann Arbor, MI, US.

Conflict of Interest Disclosures

None reported.