By: Gali Halevi, Levy Library, Icahn School of Medicine at Mount Sinai, email@example.com
Judit Bar-Ilan, Department of Information Science, Bar-Ilan University, Judit.Bar-Ilan@biu.ac.il
The process of scientific publishing is comprised of several stages, one of which is a review of the literature. In order to establish the novelty of scientific discoveries and to contextualize results, a literature review is included in each publication. Prior to publishing a new article, researchers must be familiar with prior works upon which they establish their new contribution. Usage behavior has been studied throughout the years and correlations were found between usage and citations. In order to gain an in-depth understanding of how scientists use journals it is becoming important to also consider which journals they publish in and whether these two factors correlate. Usage can be measured by documents’ views, downloads and shares on reference managers, social media platforms and other metrics. Usage and publications correlations can inform researchers’ information behavior, collection development and trends in emerging topics and areas.
Who We Studied
In this study we examined Mount Sinai researchers encompassing the Icahn School of Medicine at Mount Sinai and the Graduate School of Biomedical Sciences. Mount Sinai’s multidisciplinary Institutes have 34 academic departments focusing on collaborative research, clinical activities, and education. The Mount Sinai Health System includes more than 7,000 physicians and scientists and over 600 postdoctoral fellows.
How We Conducted Our Study
All of Mount Sinai research publications studies were associated with “The Icahn School of Medicine at Mount Sinai” as the affiliation. Using Scopus, we searched for “The Icahn School of Medicine” in the affiliation search field. Overall, Scopus retrieved 3,052 documents assigned to “Icahn School of Medicine” scientists. We limited the search to only articles, which resulted in 2,260 publications in 2015. We used the “analyze results” function in Scopus which allows downloading the list of top journals and subject areas in the set. Overall, Scopus retrieved 160 journals in which Icahn School of Medicine scientists published and 1072 articles in 2015. Using Serial Solutions, a platform which tracks usage per journal, we retrieved usage counts per each of the 160 journals. 2015 usage for 144 journals was tracked by Serial Solutions and included in our dataset. Since Serial Solutions does not differentiate between HTML views and PDF downloads, “usage” in this study is referred to as the overall views/downloads per each journal.
Our datasets, therefore, top 31 journals, 971 articles published by Icahn School of Medicine scientists and 730,989 downloads/views of articles of these journals in 2015.
What Our Results Show
Our results show that when comparing the two sets of ‘highest journals published in” and “highest journals used.” There are only 7 titles that appear in both. As can be seen in Figure 1, highly used journals are not necessarily the most published in and vice versa.
Figure 1 Top-journals both in publications and in usage
There was no Spearman correlation between the number of articles published in a journal and it’s JIF (Journal Impact Factor). The lowest JIF of the papers with 10 or more articles is 3.234, and the highest 55.873, where the weighted average is 10.02. There was no significant Spearman correlation between usage and JIFs either, but the r was at least somewhat meaningful (r=.401, p>.099). See figure 2.
Figure 2: correlations between JIF and Number of publications
The data also shows that 37% of the articles in our dataset were published in the top journals by Mount Sinai researchers. Overall, the weighted average JIFs of the top-used journals is much higher on average (33.88) than the weighted average JIF of the top-journals published in (10.02).
Conclusion: No Correlation
In our case, we could not find correlations between the journals that are most used to the ones that are most published in. Out of the 31 unique titles, only 7 could be found in both sets of highly used and highly published in journals. We also could not find correlations between the numbers of articles published in specific journals and their JIF, or a correlation between the highly used journals and their JIF. Therefore, it will be difficult to use this method as a collection development tool without further insight into publications and usage selections.
While most scientists aim to publish in high JIF journals, the motivation behind their reading selections and the relationship between the journals they read and the journals they publish in is still unclear. Further research should include interviews with top published scientists in order to better understand the relationship between their reading selection and their selection of journals they publish in.
http://www.creatomatic.co.uk. (n.d.). About. Retrieved from https://www.projectcounter.org/about/
Maflahi, N., & Thelwall, M. (2016). When are readership counts as useful as citation counts? Scopus versus Mendeley for LIS journals. Journal of the Association for Information Science and Technology, 67(1), 191–199.in publications and in usage