Coauthors' contributions to major papers published in the AJR: frequency of undeserved coauthorship

Over half of the major papers published in the American Journal of Roentgenology (AJR) have five or more coauthors. This project was designed to evaluate the specific contributions of coauthors and the prevalence of undeserved authorship in major papers from institutions in the United States.

Questionnaires were mailed to the first author of 275 major papers from institutions in the United States that were published in the AJR in 1992 and 1993. Questions focused on coauthors' contributions to research design, data collection, data analysis, and manuscript preparation, and on undeserving authorship.

One hundred ninety-six (72%) of the surveys were returned. Ninety-nine percent of first authors, 75% of second authors, fewer than half of third authors, and one third of fourth authors and beyond were said to have contributed to at least three of the following: research design, data collection, data analysis, and manuscript preparation (p < .02). A strong correlation was indicated between authorship position and contribution (r = -.69, p < .001), with a mean overall contribution of 63 +/- 17% (mean +/- SD) for the first author, 20 +/- 12% for the second author, 10 +/- 7% for the third author, 7 +/- 6% for the fourth author, and 5 +/- 5% for all other authors. Coauthors were listed in decreasing order of contribution in 70% of articles. However, the last author was the second major contributor in 10% of articles with three or more authors. The incidence of "undeserved" coauthors increased from 9% on papers with three authors to 30% on papers with more than six authors (mean, 17%; r = .97; p < .001). Undeserved authorship was attributed largely to individuals who contributed only cases (29%) or who created a sense of obligation or fear in the first author (40%). Manuscripts were more likely to include an undeserved coauthor when the first author was a nontenured staff member (45%) than when he or she was tenured faculty (28%) (p < .02). When decision about authorship were made at project conception, there were fewer coauthors (3.9 versus 5.4, p < .02) and a lower incidence of manuscripts with undeserving coauthors (23% versus 47%, p < .01). The final manuscript was read by all coauthors in 80% of manuscripts, and all coauthors were thought to understand the manuscript to the extent they could publicly defend it in 78% of manuscripts. The most commonly cited reason that otherwise honest individuals accept undeserved authorship was academic promotion.

Undeserved authorship is a common and serious problem that is motivated primarily by academic promotion policies. The first two authors are said to account for the preponderance of work in almost all major papers.