Conjunctive Analysis of Case Configurations: An Exploratory Method for Discrete Multivariate Analyses of Crime Data

• NCJ Number: 222613
• Journal: Journal of Quantitative Criminology Volume: 24 Issue: 2 Dated: June 2008 Pages: 227-241
• Author(s): Terance D. Miethe; Timothy C. Hart; Wendy C. Regoeczi
• Date Published: June 2008
• Length: 15 pages
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

Drawing on existing methods for discrete multivariate analysis, this study describes an alternative technique for exploring causal relationships among categorical variables, which is called "conjunctive analysis of case configurations"; it is illustrated in the study of Federal sentencing of drug offenders.

Abstract

The conjunctive analysis of case configurations is a simple method of discrete multivariate analysis that can be easily applied to both exploratory and confirmatory research. This method provides a middle ground between the focus on specificity and multiple causality that underlies most qualitative research and the variable-oriented search for general patterns across contexts in most quantitative research. Even for researchers who prefer variable-oriented quantitative methods, conjunctive analysis can augment and inform their substantive analysis by identifying possible problems with multicollinearity and low-cell frequencies. In illustrating conjunctive analysis, the author examined the Federal sentencing of 1,358 drug offenders from 1997 through 1998. The independent variables include the type of offense (drug trafficking, possession, and other drug offenses); the offender's prior record (prior arrest record, no prior arrests); gender; and race. The dependent variable is whether or not the offender received a prison sentence. Under the Federal sentencing guidelines, 92 percent of the drug offenders in this sample received a prison sentence. One basic way of using conjunctive analysis in assessing the main effects and interaction effects of particular variables involves the examination of the particular characteristics of drug cases that are associated with the lowest and highest risks of imprisonment. This approach involves two basic steps: rank the case configuration according to the relative risks of imprisonment within them and compare the relative prevalence of particular categories of each variable among the highest and lowest ranked groups of case configuration. 5 tables, 25 references, and appended software syntax for conjunctive analysis