A Sociological Approach to Fraud in Science

This paper extends sociological theory into the controversial area of the presentation of fraudulent scientific results. Although representative evidence does not exist, sufficient cases are known to reveal that fraud is not uncommon in science. Using a ’systems ’ approach first, areas appear within science and on its periphery where the scientific community’s sanctions are inoperative. On the margins of science, these categories include ’hired hands’, pseudo-scientists and industrial scientists. Within the scientific community the lack of guaranteed detection of fraud also creates circumstances where sanctions are inapplicable. Within this framework, the ’action ’ approach indicates possible factors motivating individuals toward fraud. For many groupings, a major factor appear to be a commitment to a theory so strong that it overrides disconfirmation, and diminishes fears of detection. Other factors include career considerations and political views. In each case evidence is produced to show, prima facie, that fraud is prevalent in the circumstances outlined. There is now a good deal of evidence that fraud in science is not uncommon, and exists in many parts of the scientific community. Public concern is indicated by a series of recent publicised cases (Broad, 1981) and a United States congressional investigation (Walton, 1981). In this paper, theoretical perspectives upon fraud in science are developed and predictions are made concerning its distribution. Some writers-such as Manwell and Baker (1981) have defined fraud, or cheating in science fairly broadly, including such behaviour as appropriating another’s work and excluding students from authorship in work for which the latter were responsible. However, for the purposes of this paper fraud is defined as the presentation of scientific results which have either been modified (’data massage’), or obtained from a ’rigged ’ experiment, or, in some cases, the presentation of ’results ’ that are completely imaginary. As Jones (1980; 48) points out, there is near unanimity in the scientifi