Coauthorship in physics

In a large and detailed survey on the ethics of scientific coauthorship, members of the American Physical Society (APS) were asked to judge the number of appropriate coauthors on his or her last published paper. Results show that the first or second coauthors are more appropriate than later coauthors about whom there is equal and considerable doubt. The probability of any third and subsequent coauthors being judged as inappropriate is 23% for the APS guideline, 67% for the tighter guideline of the International Committee of Medical Journal Editors, 59% if the guideline requires "direct contributions to scientific discovery or invention". Only 3% of respondents report having personally rejected an undeserving scientist who expected to be an author on the last published paper. Respondents seem to be divided into two non-overlapping populations--those who report no inappropriate coauthorship and those who have a more graduated view.