An analysis of citation functions in the humanities and social sciences research from the perspective of problematic citation analysis assumptions
Chi-Shiou Lin
DOI: https://doi.org/10.1007/s11192-018-2770-2
IF: 3.801
2018-05-11
Scientometrics
Abstract:Citation analyses are based on two questionable assumptions that significantly influence the value of citation count: the equal contribution assumption (i.e., each citation contributes equally to the citing paper) and the positive endorsement assumption (i.e., each citation is viewed as an endorsement from the citing paper to the cited work). This study employed a citation context analysis approach to examine the distributions of essential versus perfunctory citations as well as confirmative versus negational citations in humanities and social sciences (H&SS) research articles to determine to what extent the two problematic assumptions affect the validity of citation analyses in those fields. The sample comprises 360 articles representing six subject areas: Chinese literature, history, arts (i.e., the humanities subjects); sociology, economics, and psychology (i.e., the social sciences subjects). Excluding citations to primary sources—e.g., manuscripts, archival materials, monographs that were the subjects of study rather than information sources, a total of 25,617 in-text bibliographic citations were analyzed. The findings show that the distributions of citation functions for those six subject fields were all significantly different, meaning that disciplinary differences exist in H&SS citations. Also, for the H&SS fields, perfunctory citations may have inflated the contribution of the cited works much more than negational citations.
information science & library science,computer science, interdisciplinary applications