%0 Journal Article %1 leydesdorff2014citation %A Leydesdorff, Loet %A Milojević, Staša %D 2014 %I SSRN %J SSRN eLibrary %K BRD impact_factor soziologie zeitschrift %R 10.2139/ssrn.2541908 %T The Citation Impact of German Sociology Journals: Some Problems with the Use of Scientometric Indicators in Journal and Research Evaluations %X We analyze the citation impacts (‘cited’) and citation practices (‘citing’) of sociological journals which publish mainly in German, and discuss major drawbacks of using the journal impact factor (IF) to assess the quality of these journals. First, sociological literature moves very slowly in terms of citations, whereas journal impact factors (IF) measure short-term impact at research fronts. Second, the citation distributions are heavily skewed because of the so-called Matthew effect (Merton) of cumulative advantages; one should not use quasi-averages (such as the IF) given this skewness. The alternative of using non-parametric statistics (e.g., percentiles), however, requires delineation of the reference sets. We discuss the unsolved problems in the case of (inter)disciplinary delineations and show empirically the ecological fallacy in nevertheless attributing journal characteristics to individual papers. Algorithmic constructs (e.g., the various rankings) cannot be used for policy or management purposes without validation or specification of statistical error.

@article{leydesdorff2014citation, abstract = {We analyze the citation impacts (‘cited’) and citation practices (‘citing’) of sociological journals which publish mainly in German, and discuss major drawbacks of using the journal impact factor (IF) to assess the quality of these journals. First, sociological literature moves very slowly in terms of citations, whereas journal impact factors (IF) measure short-term impact at research fronts. Second, the citation distributions are heavily skewed because of the so-called Matthew effect (Merton) of cumulative advantages; one should not use quasi-averages (such as the IF) given this skewness. The alternative of using non-parametric statistics (e.g., percentiles), however, requires delineation of the reference sets. We discuss the unsolved problems in the case of (inter)disciplinary delineations and show empirically the ecological fallacy in nevertheless attributing journal characteristics to individual papers. Algorithmic constructs (e.g., the various rankings) cannot be used for policy or management purposes without validation or specification of statistical error.}, added-at = {2014-12-28T17:05:19.000+0100}, author = {Leydesdorff, Loet and Milojević, Staša}, biburl = {https://www.bibsonomy.org/bibtex/26ad67d85a48d27e34cc59e76de05feae/wdees}, description = {The Citation Impact of German Sociology Journals: Some Problems with the Use of Scientometric Indicators in Journal and Research Evaluations by Loet Leydesdorff, Staša Milojević :: SSRN}, doi = {10.2139/ssrn.2541908}, interhash = {13e49e29296e0fcb2481a20bbe974328}, intrahash = {6ad67d85a48d27e34cc59e76de05feae}, journal = {SSRN eLibrary}, keywords = {BRD impact_factor soziologie zeitschrift}, language = {English}, location = {http://ssrn.com/paper=2541908}, publisher = {SSRN}, timestamp = {2014-12-28T17:05:19.000+0100}, title = {The Citation Impact of German Sociology Journals: Some Problems with the Use of Scientometric Indicators in Journal and Research Evaluations}, type = {Working Paper Series}, year = 2014 }