Early Statistical Findings and Authorship Misattribution: An Unsystematic Review of the Literature

Abstract

39 For historical accuracy and in fairness to the authors of original work, it is right and proper to acknowledge their contributions to scientific advances. I first encountered authorship misattribution in the statistical literature in the early 1980s. In 1971, Richard Light reported an apparently new measure of observer agreement in Psychological Bulletin known as the conditional kappa coefficient. Shortly afterward, I came across a copy of Fingerprints (Galton. 1892), in which Francis Galton described a “centesimal scale” that he used to assess the similarity of the patterns on corresponding fingers of the left and right hands. Comparison with Light’s paper shows that Galton had proposed the same measure, which I pointed out in a letter to Biometrics in 1985. Since then, I have seen other examples of author misattribution in a statistical context. These have been reported as individual cases, but I have yet to discover a collection of reports of misattribution obtained through a literature review. This article provides a way to rectify that situation.