Search

It is now widely believed that the more numerous the authors of a scientific paper, the greater is the likelihood of higher citation impact. By contrast, this paper considers a set of single- and multiple-authored publications in a group of matched journals that have resulted from collaborative research networks funded by the Austrian Science Fund, and presents evidence that no statistically significant relationship is found between multiple-authored papers and higher citation impact over single-authored papers. Moreover, within the data set examined, some evidence is found of a negative relationship between increasing numbers of authors and higher citation impact. The implication is drawn that where research is carried out within larger networks where researchers may benefit from a more general rather than a more specific collaboration, some researchers may publish their more important work through single-authored papers in order to enhance their reputations. Further implications of these findings are then considered for research funders and researchers. [ABSTRACT FROM AUTHOR]

Starting out from a broad concept of research output, this article looks at the question as to what research outputs can typically be expected from certain disciplines. Based on a secondary analysis of data from final project reports (ex post research evaluation) at the Austrian Science Fund (FWF), Austria’s central funding organization for basic research, the goals are (1) to find, across all scientific disciplines, types of funded research projects with similar research output profiles; and (2) to classify the scientific disciplines in homogeneous segments bottom-up according to the frequency distribution of these research output profiles. The data comprised 1,742 completed, FWF-funded research projects across 22 scientific disciplines. The multilevel latent class (LC) analysis produced four LCs or types of research output profiles: ‘Not Book’, ‘Book and Non-Reviewed Journal Article’, ‘Multiple Outputs’, and ‘Journal Article, Conference Contribution, and Career Development’. The class membership can be predicted by three covariates: project duration, requested grant sum, and project head’s age. In addition, five segments of disciplines can be distinguished: ‘Life Sciences and Medicine’, ‘Social Sciences/Arts and Humanities’, ‘Formal Sciences’, ‘Technical Sciences’, and ‘Physical Sciences’. In ‘Social Sciences/Arts and Humanities’ almost all projects are of the type ‘Book and Non-Reviewed Journal Article’, but, vice versa, not all projects of the ‘Book and Non-reviewed Journal Article’ type are in the ‘Social Sciences/Arts and Humanities’ segment. The research projects differ not only qualitatively in their output profile; they also differ quantitatively, so that projects can be ranked according to amount of output. [ABSTRACT FROM AUTHOR]