F1000Prime recommended articles and their citations: an exploratory study of four journals

This study examined F1000Prime recommended research and review articles published in Cell, JAMA: The Journal of the American Medical Association, The Lancet, and The New England Journal of Medicine (NEJM) in 2010. The analyses included (1) the classifications assigned to the articles; (2) differences in Web of Science (WoS) citation counts over 9 years between the articles with F1000Prime recommendations and the other articles of the same journal; (3) correlations between the F1000Prime rating scores and WoS citation counts; (4) scaled graphic comparisons of the two measures; (5) content analysis of the top 5 WoS cited and top 5 F1000Prime scored NEJM articles. The results show that most of the recommended articles were classified as New Finding, Clinical Trial, Conformation, Interesting Hypothesis, and Technical Advance. The top classifications differred between the medical journals (JAMA, The Lancet, and NEJM) and the biology journal (Cell); for the latter, both New Finding and Interesting Hypothesis occurred more frequently than the three medical journals. The articles recommended by F1000 Faculty members were cited significantly more than other articles of the same journal for the three medical journals, but no significance was found between the two sets of articles in Cell. The correlations between the F1000Prime rating scores and WoS citation counts of the articles in the same journal were significant for the two medical journals (The Lancet and NEJM) and the biology journal (Cell). NEJM showed significances in both the upper quantile (top 50%), and the upper quartile (top 25%) sets. One of the medical journals, JAMA, did not show any significant correlation between the two measures. Despite the significant correlations of the three journals, Min–Max scaled graphic comparisons of the two measures did not reveal any patterns for predicting citation trends by F1000Prime rating scores. The peak citation year of the articles ranged from 2 to 8 years after the publication year for NEJM. Content analysis of the top-cited and top-scored NEJM articles found that highly commendable papers with comments such as “exceptional,” “landmark study,” or “paradigm shift” received varied rating scores. In comparison, some of the results corroborate with previous studies. Further studies are suggested to include additional journals and different years as well as alternative methods. Studies are needed to understand how F1000 Faculty assign ratings and what criteria they use. In addition, it is also worth investigating how F1000Prime users perceive the meanings of the ratings.