The multidimensionality of crime: A comparison of techniques for scaling delinquent careers

The construction of typologies of criminal behavior can benefit from the use of multidimensional analytic methods. Yet while some studies have applied such techniques to crime data (e.g., Shortet al, 1963; Nutch and Bloombaum, 1968; Chaiken and Chaiken, 1982), few have examined the assumptions of these methods as they apply to arrest histories. We argue that arrest histories represent a special form of data that are not ideally suited to standard multidimensional analyses. An examination of the different theoretical assumptions of factor analysis, multidimensional scaling, and variance centroid scaling (a form of correspondence analysis) reveals marked difierences in what is being uncovered by the analysis. In general, these claims are supported by an application of each technique to the arrest histories of 767 chronic juvenile delinquents.