RESEARCH ON AUTHORSHIP

Wendela Hoen, Henk Walvoorth, and John Overbeke
Nederlands Tijdschrift voor Geneeskunde, PO Box 7591, 1070 AZ Amsterdam, The Netherlands

Objective: To investigate what criteria are used by authors of articles in the Dutch Journal of Medicine (Nederlands Tijdschrift voor Geneeskunde, NTvG) to define authorship and also the criteria according to which authors were listed.

Design: Descriptive study; survey of 450 authors of original articles published in 1995 in NTvG with 3 or more authors. This survey contained questions concerning each author's contribution to the article (eg, study design, material, practice, statistics, and writing); questions on the criteria for authorship developed by the International Committee of Medical Journal Editors (ICMJE); and open-structured questions on how they determined who was to be mentioned as an author of their article and how order of authorship was determined.

Results: In 1995, 115 original articles were published with 3 or more authors. The 450 authors of the 155 manuscripts were surveyed; 362 authors returned the survey (overall response rate, 80.4%); 352 of the surveys could be analyzed. These questionnaires were returned by 92 first authors, 83 last authors, and 117 authors in between. The questionnaire consisted of 23 items of which 7 were related to the ICMJE criteria. The first 5 most frequently positive answered questions were all ICMJE criteria. Critical reading was done by 86.1% of the authors, approval of the version to be published by 84.7%, design of the study by 74.7%, conception of the study by 64.2%, and revision by 63.4%. No contribution at all was made by 1.1%. The discrepancy between the respondent's own answers and those of the coauthors was calculated and showed a Gauss distribution. On average respondents claimed only 2 points more for themselves than their coauthors rated for them on a scale from +22 to -22. Sixty-three percent of the respondents met all 3 ICMJE criteria for authorship; 93% met 2 of the 3 criteria; and 57% of the respondents did not know about the criteria. The criteria mentioned for the order of the authors were inconsistent and differed from group to group and from faculty to faculty. For the majority of the respondents this was a main problem.

Conclusion: There is much inconsistency in criteria for authorship and even more in guidelines for the authors' order. The ICMJE authorship criteria are reasonably known and met by many authors, but they are of no help for determining order of authorship. Clearer and more practical guidelines should be developed.

Return to Session Information