The method of paired comparisons for social values

This is an attempt to apply the ideas of psychophysical measurement in the field of social values. Some of the psychophysical methods have been applied in a crude way to the measurement of educational products such as handwriting and English composition, and it seems feasible to apply the same ideas as well to social values although the attempt cannot readily be made without making compromises that the psychophysicist would not tolerate. The application of the principles of psychophysical measurement to educational products has been made with more or less similar logical handicaps but these do not seem to have disturbed the popularity of these methods in the field of educational measurement. Since the final results show a rather satisfactory internal consistency, one may possibly assume that the methods, the theoretically imperfect, have some value also in social psychology. For the present experiment the seriousness of different crimes or offenses was chosen for such measurement. The seriousness of an offense we shall assume to be the seriousness as judged rather than as measured in terms of objective consequences or in some normative way. Nineteen offenses were judged: abortion, adultery, arson, assault and battery, bootlegging, burglary, counterfeiting, embezzlement, forgery, homicide, kidnapping, larceny, libel, perjury, rape, receiving stolen goods, seduction, smuggling, and vagrancy. The offenses were arranged in pairs so that every one of them was paired with every other one. The total number of pairs of offenses presented was therefore n(n-1)/2 = 171. College student subjects were presented with the paired offenses and asked to choose which they considered to be the more serious crime. Results show that qualitative judgments of a rather intangible sort, loaded usually with personal opinion, bias, and even strong feeling, and regarded generally as the direct antithesis of quantitative measurement, are nevertheless amenable to the type of quantitative analysis which is associated historically with psychophysics. It is of some interest to see that a set of numerical values can be established by which the 171 observed proportions of judgments about crimes and offenses can be summarized in generalized form. (PsycINFO Database Record (c) 2006 APA, all rights reserved)