A Study of Factors Related to Readership of Scientific Articles

This report analysed the relationship between the number of Twitter and Mendeley views of scientific articles and the following factors: the GDP per capita of the country where the views came from, the title length, the publisher, the journal, and the subject of the article, in order to determine which factors affected how many views scientific articles accumulate. The purpose of this report was to help future researchers gain the most views possible on their published articles.

The data provided by Altmetric containing 550,000 json files, was extracted and displayed into several data frames, each one containing a different variable. The variables were then compared to views articles gained using a combination of bubble graphs, linear regression models, and scatter plots. From the results gathered, it could be seen that although the title length, the GDP per capita, and the subject of an article all affected the total views an article amassed, the journal and publisher had no notable effect on the views. In particular, title length within 41-100 characters had the most noticeable effect on the number of readers per article. Subjects, while showing that the more articles published about them brought in more total views, did not have a strong effect on the average number of viewers per article.