The relationship between journal use in a medical library and citation use

The purpose of this study was to investigate the relationship between the library journal use and journal citation use in the medical sciences. The journal use study was conducted in the Library of Veterans General Hospital--Taipei for six months. The data of citation frequency, impact factor, citation age and citation half-life were obtained from Journal Citation Reports (JCR) 1993 microfiche edition. Only 835 journals which were in JCR and in the library were compared. The correlation between use frequency and citation frequency and between use frequency and impact factor was determined based on the Spearman rank and Pearson correlation tests. The distribution comparison of use and citation age distributions was examined by using Kolmogorov-Smirnov goodness of fit tests. The t-test was used to test the difference of use half-life and citation half-life. Comparisons were also made for the four subject categories, namely, (A) clinical medicine journals; (B) life science journals; (C) hybrid journals publishing both clinical medicine and life science articles, and (D) those journals publishing neither clinical medicine nor life science articles. The results of the study showed that there is a significant correlation between use frequency and citation frequency, and between use frequency and impact factor for all titles. A significant correlation also exists between use frequency and citation frequency, and between use frequency and impact factor for journals of category A, B, and C. However, the correlation is not significant for category D journals. The mean use half-life of the total 835 journals is 3.43 years, which is significantly shorter than the citation half-life, 6.28 years. The mean citation half-life is 5.92 years if 74 journals with a half-life equal to or longer than 10 years are excluded. There is also a significant difference between the mean citation half-life and the mean use half-life for journals of each category. The journals irrelevant to either clinical medicine or life science have the largest difference in mean citation half-life and mean use half-life. The use age distributions for the mean of all the journals or of each subject category illustrate an exponentially decaying curve. Age one journals contribute the maximum use frequency. On the other hand, the citation age distributions for the mean of all the journals or of each category show a sharp initial rise from age one to three or four and then fall off in a sort of exponential decay. The age of maximum citation is typically 3 years. Comparison of use and citation age distribution in terms of percentage indicates there is a dramatic difference for age one; the two curves intersect at an age between two and three; after that the citation age curve exceeds the one for use age and is nearly parallel to the use age curve. The Kolmogorov-Smirnov two sample test indicates that the use age distribution does not fit the citation age distribution for the mean of all the journals, or journals of each subject category, except category D, or each of the top ten most used journals. The category D journals are those dedicated neither to clinical medicine nor to life science.