by Donald W. Katzner
The purpose of this book is to demonstrate that it is possible to do meaningful, significant, and sophisticated analysis in social science when the variables under consideration are, given present knowledge, incapable of measurement. No effort to 'measure' the unmeasurable is attempted. Rather, techniques for model building, such as the construction of simultaneous and periodic relation systems that do not require the existence of measures are explored. In addition to presenting a methodology enabling the investigator to deal with the unmeasured, many examples are provided that illustrate how those methods may actually be used. In addition, the book addresses the following: Where has the overwhelming focus on the quantitative (often to the exclusion of the unmeasurable or qualitative) in social science in particular, and in modern societies in general, come from? How can the use of the formalizations of model building, both in the presence and absence of measurement, be justified in social science?
What are the dangers of using proxy variables in general in the construction of models, and what are the dangers of treating variables that are only ordinally gauged as if they were cardinally or intervally measured? Finally, when only ordinal calibrations of some variables are available, what analytical methods may legitimately be employed to deal with them?