Original Paper Ophthalmic Res 2014;52:136–140 DOI: 10.1159/000366284
Received: April 29, 2014 Accepted after revision: July 30, 2014 Published online: October 4, 2014
Self-Citation Rate and Impact Factor in Ophthalmology Michael Mimouni Ori Segal Department of Ophthalmology, Meir Medical Center, Kfar Saba, and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
Key Words Impact factor · Self-citation · Self-citation rate
journal does not correlate with its SCR. Self-citation is found more often in journals with a low corrected IF and is inversely correlated with IF in the bottom half. © 2014 S. Karger AG, Basel
Abstract Purpose: To analyze the self-citation rate (SCR) of ophthalmology journals, determine its possible effect on a journal’s impact factor (IF) and compare the SCR of subspecialty journals versus general ophthalmology journals. Method: A retrospective consecutive study of ophthalmology journals listed in the Journal Citations Report (JCR) 2013. We retrieved these parameters from each journal’s report: IF, total citations, self-citations, SCR and IF without self-citations (corrected IF). Results: A significant correlation was detected between the number of self-citations and publications (R2 = 86.3, p = 0.000). Subspecialty journals had a significantly higher SCR than general journals (p = 0.017). No significant difference was found in terms of IF and corrected IF between general and subspecialty journals (p = 0.260 and p = 0.108, respectively). No significant correlation between IF and SCR was detected (p = 0.099). The corrected IF was inversely correlated with SCR (R2 = –32.6, p = 0.013). An inverse correlation was detected between SCR and IF in the 29 ophthalmology journals with the lowest IF (R2 = –57.3, p = 0.001). Conclusion: Unlike other fields of medicine, the IF of an ophthalmology
© 2014 S. Karger AG, Basel 0030–3747/14/0523–0136$39.50/0 E-Mail [email protected] www.karger.com/ore
Introduction
In 1955 Garfield coined the term ‘impact factor’ (IF) [1]. He described the need to evaluate the significance of a particular work and its impact on scientific literature. This led to the production of the Science Citation Index [2] which served as a foundation for the Journal Citation Reports® (JCR) [3]. The IF of a given journal is the average number of citations received per paper published during the 2 previous years [4]. Currently, the IF is a common method for assessing the quality of a scientific journal, and many journals state their IF and ranking on their websites [4–6]. However, the IF has been criticized for having many flaws, and alternatives have been proposed [7–12]. Self-citation, as defined by the JCR, occurs when an article in a journal makes reference to another article
M. Mimouni and O. Segal contributed equally to this study.
Michael Mimouni, MD Department of Ophthalmology Meir Medical Center 59 Tchernichovsky St., Kfar Saba 44410 (Israel) E-Mail michael @ intername.co.il
within the same journal [13]. Thus, the IF of a journal can be ‘improved’ by increasing the amount of self-citations. The self-citation rate (SCR) has been defined by the Institute for Scientific Information as the ratio of journal selfcitations to the number of times it is cited by all other journals, including itself [14]. The SCR of a journal has been shown to correlate with its IF in various fields of medicine, such as anesthesia [15], orthopedics [16] and gastroenterology and hepatology [17]. In contrast, a similar correlation was not found in otolaryngology [18]. The aim of this study was to: (1) analyze the SCR of ophthalmology journals registered in the JCR; (2) determine the possible effect of the SCR of a given journal on its IF; (3) compare the SCR of subspecialty journals versus general ophthalmology journals and (4) determine whether the IF of a journal after removal of self-citations (corrected IF) is still correlated with the SCR.
Methods The 2013 JCR was accessed, and a list of all of the ophthalmology journals was retrieved. Each journal’s full JCR report was analyzed, and the following parameters were retrieved: IF, total publications used to calculate IF, total citations used to calculate IF, self-citations used to calculate IF, SCR and IF without self-citations (corrected IF). Each journal was independently reviewed by 2 separate investigators and categorized as either being a general ophthalmology journal or a subspecialty journal. A subspecialty journal was defined as one which declared in its ‘aims and scopes’ that its main focus was on a specific area or subject within ophthalmology. Journals that did not offer a full text version of their papers in English were excluded from the subspecialty analyses. Minitab version 16.1 (State College, Pa., USA) was used for statistical analyses. Data are presented as means ± SD or medians (with ranges) as appropriate for continuous variables or as numbers (with percentages) for categorical variables. We used the Kolmogorov-Smirnov test to identify normality of distributions in groups and found a nonnormal distribution in both SCR and IF. Thus, nonparametric tests (Kruskal-Wallis or Spearman’s rank correlation) were used for analyses, as needed. For all tests, p values
Received: April 29, 2014 Accepted after revision: July 30, 2014 Published online: October 4, 2014
Self-Citation Rate and Impact Factor in Ophthalmology Michael Mimouni Ori Segal Department of Ophthalmology, Meir Medical Center, Kfar Saba, and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
Key Words Impact factor · Self-citation · Self-citation rate
journal does not correlate with its SCR. Self-citation is found more often in journals with a low corrected IF and is inversely correlated with IF in the bottom half. © 2014 S. Karger AG, Basel
Abstract Purpose: To analyze the self-citation rate (SCR) of ophthalmology journals, determine its possible effect on a journal’s impact factor (IF) and compare the SCR of subspecialty journals versus general ophthalmology journals. Method: A retrospective consecutive study of ophthalmology journals listed in the Journal Citations Report (JCR) 2013. We retrieved these parameters from each journal’s report: IF, total citations, self-citations, SCR and IF without self-citations (corrected IF). Results: A significant correlation was detected between the number of self-citations and publications (R2 = 86.3, p = 0.000). Subspecialty journals had a significantly higher SCR than general journals (p = 0.017). No significant difference was found in terms of IF and corrected IF between general and subspecialty journals (p = 0.260 and p = 0.108, respectively). No significant correlation between IF and SCR was detected (p = 0.099). The corrected IF was inversely correlated with SCR (R2 = –32.6, p = 0.013). An inverse correlation was detected between SCR and IF in the 29 ophthalmology journals with the lowest IF (R2 = –57.3, p = 0.001). Conclusion: Unlike other fields of medicine, the IF of an ophthalmology
© 2014 S. Karger AG, Basel 0030–3747/14/0523–0136$39.50/0 E-Mail [email protected] www.karger.com/ore
Introduction
In 1955 Garfield coined the term ‘impact factor’ (IF) [1]. He described the need to evaluate the significance of a particular work and its impact on scientific literature. This led to the production of the Science Citation Index [2] which served as a foundation for the Journal Citation Reports® (JCR) [3]. The IF of a given journal is the average number of citations received per paper published during the 2 previous years [4]. Currently, the IF is a common method for assessing the quality of a scientific journal, and many journals state their IF and ranking on their websites [4–6]. However, the IF has been criticized for having many flaws, and alternatives have been proposed [7–12]. Self-citation, as defined by the JCR, occurs when an article in a journal makes reference to another article
M. Mimouni and O. Segal contributed equally to this study.
Michael Mimouni, MD Department of Ophthalmology Meir Medical Center 59 Tchernichovsky St., Kfar Saba 44410 (Israel) E-Mail michael @ intername.co.il
within the same journal [13]. Thus, the IF of a journal can be ‘improved’ by increasing the amount of self-citations. The self-citation rate (SCR) has been defined by the Institute for Scientific Information as the ratio of journal selfcitations to the number of times it is cited by all other journals, including itself [14]. The SCR of a journal has been shown to correlate with its IF in various fields of medicine, such as anesthesia [15], orthopedics [16] and gastroenterology and hepatology [17]. In contrast, a similar correlation was not found in otolaryngology [18]. The aim of this study was to: (1) analyze the SCR of ophthalmology journals registered in the JCR; (2) determine the possible effect of the SCR of a given journal on its IF; (3) compare the SCR of subspecialty journals versus general ophthalmology journals and (4) determine whether the IF of a journal after removal of self-citations (corrected IF) is still correlated with the SCR.
Methods The 2013 JCR was accessed, and a list of all of the ophthalmology journals was retrieved. Each journal’s full JCR report was analyzed, and the following parameters were retrieved: IF, total publications used to calculate IF, total citations used to calculate IF, self-citations used to calculate IF, SCR and IF without self-citations (corrected IF). Each journal was independently reviewed by 2 separate investigators and categorized as either being a general ophthalmology journal or a subspecialty journal. A subspecialty journal was defined as one which declared in its ‘aims and scopes’ that its main focus was on a specific area or subject within ophthalmology. Journals that did not offer a full text version of their papers in English were excluded from the subspecialty analyses. Minitab version 16.1 (State College, Pa., USA) was used for statistical analyses. Data are presented as means ± SD or medians (with ranges) as appropriate for continuous variables or as numbers (with percentages) for categorical variables. We used the Kolmogorov-Smirnov test to identify normality of distributions in groups and found a nonnormal distribution in both SCR and IF. Thus, nonparametric tests (Kruskal-Wallis or Spearman’s rank correlation) were used for analyses, as needed. For all tests, p values
