What Citation Data Showed Across Nine Sport Journals
A look at 15 years of publications and citations in nine sport journals, what we found about visibility, and what we would do differently.

Between 2006 and 2020, we analysed publications and citations in nine sport economics, management, and sociology journals. We wanted to see whether women's work was cited at similar rates to men's and where gaps might appear. The data showed a persistent gap in average citations per article, even after accounting for journal and year. But the gap narrowed in journals with more women on the editorial board. We also found that articles with women as first authors were less likely to be cited in review articles. These patterns matter for visibility, and they shaped how we designed our later workshops. You can read more about reading citation counts responsibly.
Why we looked at citation data
We started this project in early 2021 after a mentoring session where a junior scholar asked why her well-received paper seemed to disappear from conversations. We did not have a good answer. So we decided to look at the numbers ourselves. We chose nine journals that publish sport economics, sport management, and sport sociology work, and we pulled every article from 2006 to 2020. That gave us 4,812 articles and their citation counts as of mid-2022. We focused on first-author gender, which we coded from names and, where needed, from public profiles. We know that method is imperfect, and we say so up front.
The goal was not to rank journals or scholars. It was to understand whether visibility, measured by citations, differed by gender and what factors might explain it. This report summarises what we found and what we plan to change in our next round.
The work built on our earlier workshop on planning a workshop that works for early-career women, where participants asked for concrete data on publication visibility.
What the data showed
Across all nine journals, articles with women as first authors received about 18% fewer citations on average than articles with men as first authors. The gap was not uniform. In journals with at least 40% women on the editorial board, the gap shrank to about 7%. In journals with fewer than 20% women on the board, the gap was closer to 25%. We also looked at review articles: articles by women first authors were cited in reviews about 12% less often than articles by men first authors, even when published in the same journal and year.
We checked whether the gap could be explained by article type, length, or topic area. None of those factors fully accounted for it. The pattern held across subfields, though it was slightly larger in sport economics than in sport sociology.
| Measure | Result |
|---|---|
| Articles analysed | 4,812 |
| Journals included | 9 |
| Years covered | 2006–2020 |
| Average citation gap | 18% |
| Gap in journals with ≥40% women on editorial board | 7% |
| Gap in journals with <20% women on editorial board | 25% |
| Review-article citation gap | 12% |
We are careful not to overstate these numbers. They describe patterns in our sample, not proof of causation. But they are consistent with what many colleagues have described anecdotally.
What worked and what did not
The coding process worked well. Two researchers coded independently, and disagreements were resolved by a third. We achieved 94% agreement on first-author gender. The citation counts from the database we used were reliable and updated regularly. We also benefited from advice from colleagues who had done similar work in other fields.
What did not work as well: our journal selection was limited to journals we could access easily. We missed several regional journals and several open-access titles that might show different patterns. Our gender coding relied on names, which may misclassify some authors, especially those with gender-neutral names or non-Western naming conventions. We did not track self-citations, which could matter for visibility. And we did not look at co-authorship networks, which likely play a role.
What we would do differently
Next time, we would expand to at least 15 journals, including open-access and regional titles. We would collect self-citation data and co-authorship information. We would also involve a larger coding team with expertise in diverse naming conventions. And we would publish our coding protocol openly so others can replicate or challenge our findings. These changes would make the results more robust and more useful for the workshops we run.
We also learned that timing matters. Our data collection took six months, longer than planned. For future rounds, we will build in more time and share preliminary findings earlier with participants, so the research feeds directly into mentoring conversations.
The findings have already shaped our mentoring sessions. In our first mentoring pairings, we now talk about citation patterns and how to make work more visible. We also encourage mentees to consider journals with diverse editorial boards, not just prestige.
What this means for visibility
The data suggest that visibility is not just about the quality of the work. Structural factors, like who sits on editorial boards and who gets cited in reviews, seem to matter. For individual scholars, that means being strategic about where to publish and how to promote work. For journals and associations, it means looking at editorial board composition and review practices. We are not the first to say this, but our data add local evidence from sport academia.
We will continue to share these findings in workshops and at conferences. If you want to dig into the numbers, the full table is above. For a broader discussion of why this matters, see publication visibility matters for women in sport academia.
We are grateful to the 14 workshop participants who gave feedback on early versions of this analysis. Their questions pushed us to be clearer about what the data can and cannot tell us.
Questions people ask
What journals did you analyse?
We looked at nine journals in sport economics, sport management, and sport sociology, covering articles published from 2006 to 2020. We did not include open-access or regional titles in this round.
How did you determine first-author gender?
We coded gender from names and, where needed, from public profiles. Two researchers coded independently and reached 94% agreement. We recognise this method has limitations.
Will you share the full dataset?
We are preparing a public version of the coding protocol and summary tables. We will announce it on our site when it is ready. We will not share individual-level citation data.


