Misuses of Statistics in the Social Sciences: A Quantitative Approach to the Manipulation of Numbers

Christopher J. Becker, The College of Wooster

Abstract

This Independent Study examines reports of statistical misuses and tries to develop a theoretical framework for understanding what motivates people to break the rules dictating the proper usage of statistics. Existing literature on this topic is still lacking but an article by Spirer and Jaffe (1980) provides a good overview of the extent of statistical misuses. A few articles use more specific examples to illustrate this problem while other articles deal with scientific fraud in general. Three theories dealing with deviance are examined for their relevance to the understanding of individual's motives for misapplying statistics. Also, theories by Merton and Broad and Wade dealing with the structure of science are reviewed to gain a fuller comprehension of institutional forces which can pressure scientists to commit fraud. Two methods of factor analysis are examined as a means of demonstrating the intricacies involved in using statistics. Then, two ways of misusing statistics are demonstrated. Further studies are suggested, such as interviewing authors about their experiences with statistics, whether they have misused or been tempted to misuse them.