A systematic review of research on the meaning, ethics and practices of authorship across scholarly disciplines

Abstract

Background: The purpose of this systematic review was to evaluate evidence about authorship issues and provide synthesis of research on authorship across all research fields.

Methods: We searched bibliographical databases to identify articles describing empirical quantitive or qualitative research from all scholarly fields on different aspects of authorship. Search was limited to original articles and reviews.

Results: The final sample consisted of 123 articles reporting results from 118 studies. Most studies came for biomedical and health research fields and social sciences. Study design was usually a survey (53%) or descriptive study (27%); only 2 studies used randomized design. We identified four 4 general themes common to all research disciplines: authorship perceptions, definitions and practices, defining order of authors on the byline, ethical and unethical authorship practices, and authorship issues related to student/non-research personnel-supervisor collaboration. For 14 survey studies, a meta-analysis showed a pooled weighted average of 29% (95% CI 24% to 35%) researchers reporting their own or others' experience with misuse of authorship. Authorship misuse was reported more often by researcher outside of the USA and UK: 55% (95% CI 45% to 64%) for 4 studies in France, South Africa, India and Bangladesh vs. 23% (95% CI 18% to 28%) in USA/UK or international journal settings.

Interpretation: High prevalence of authorship problems may have severe impact on the integrity of the research process, just as more serious forms of research misconduct. There is a need for more methodologically rigorous studies to understand the allocation of publication credit across research disciplines.

Figure 1. Selection of the articles for the systematic review.

Search keyword was ‘authorship’, limited to article as a publication type, search performed 15 January 2010. Asterisk : inclusion criteria – quantitive or qualitative research on the definition of or criteria for authorship, authors' contribution to the research and manuscript, order of authors on the byline, opinions of researchers and/or editors on authorship criteria, opinions of researchers and/or editors on authorship order; exclusion criteria: 1. research topics which use journal articles and their authors as a starting point for studying: collaborative or citation networks; authorship in the context of citation analysis; analysis of research collaboration outputs of institutions, groups, research fields; trends in authorship in journals, groups of journals, fields, institutions, countries, geographical regions; gender of authors in journals, groups of journals, fields, institutions, countries, geographical regions; 2. analysis of authorship attribution in literature, taxonomy, and psychology/cognitive research.

Figure 2. Trends in publications on authorship research in different research areas since 1967, when the first research report was identified .

No studies were identified in humanities.

Figure 3. Forest plot of reported rates of problems with or misuse of authorship in self- or non-self reports in 14 survey studies , , , , , , , , , , , , , .

The area of a square represent sample size, horizontal lines are 95% confidence interval, diamond and vertical dotted line show the pooled weighted estimate.

Figure 4. Forest plot of reported rates of problems with or misuse of authorship in self- or non-self reports in 12 survey studies from USA, UK or international journals , , , , , , , .

The area of a square represent sample size, horizontal lines are 95% confidence interval, diamond and vertical dotted line show the pooled weighted estimate.

Figure 5. Forest plot of reported rates of problems with or misuse of authorship in self- or non-self reports in 4 survey studies from South Africa, France, India, or Bangladesh , , , .

The area of a square represent sample size, horizontal lines are 95% confidence interval, diamond and vertical dotted line show the pooled weighted estimate.