Looking Back to Move Forward

Abstract

When Journal of Research in Crime and Delinquency was first published, criminology was in the midst of an important research stream on the measurement of offending. This never solidified into a strong subdiscipline akin to psychometrics, however. After briefly discussing the goals of measurement and how they correspond to the explanation of criminal events and behavior, the authors consider how the prevailing methodological paradigm, which relies heavily on analysis of a limited number of data sets via variable-based regression techniques, may constrain measurement progress on the whole. In doing so, they highlight the imbalance between the growing sophistication of analytic models and the relative stagnation of the

Patterns and Correlates of Offender Versatility and Specialization Across a 23-Year Span for At-Risk Young Men

“…Fifth, as is the case for virtually all individual-level measures of offender specialization, the diversity index has some inherent limitations, such as the potential confound with the frequency of offending and the fact that it does not provide information on the form of specialization (e.g., Sullivan et al, 2009). Although some of these problems are alleviated by more recently applied individual-level approaches (e.g., the method proposed by Nieuwbeerta et al, 2011), these newer approaches are subject to their own set of limitations and no consensus exists about the best way to conceptualize and measure offender specialization (Nieuwbeerta et al, 2011; Sullivan et al, 2009; Sullivan & McGloin, 2014). In addition, the diversity index, which continues to be used in recent work (e.g., McGloin et al, 2007; Sullivan et al, 2009), appeared to be the best choice for the current study for the reasons described earlier.…”

“…Because of such concerns, newer individual-level techniques such as latent class analysis, multilevel item response theory modeling, and nonparametric regression have increasingly been applied to the examination of offender specialization in recent years (e.g., Baker, Metcalfe, & Jennings, 2013; Brownfield & Sorenson, 1987; Hagan, Gillis, & Brownfield, 1996; Nieuwbeerta, Blokland, Piquero, & Sweeten, 2011; Osgood & Schreck, 2007; Raudenbush, Johnson, & Sampson, 2003; Sullivan et al, 2009). To date, no consensus exists on the best way to measure specialization; each approach offers unique insights into the extent, nature, or form of offender specialization (Sullivan & McGloin, 2014; Sullivan et al, 2009).…”

Support for Balanced Juvenile Justice: Assessing Views About Youth, Rehabilitation, and Punishment

“…For the present study, an added benefit was the opportunity to increase the likelihood of full, or near full, participation in the study. Not least, use of a student population provides a cost-effective foundation on which to investigate a new measurement strategy and to test theory (Sullivan and McGloin 2014; see also Cook and Campbell 1979;Lucas 2003;Mook 1983).…”

Do Prison Administrative and Survey Data Sources Tell the Same Story? A Multitrait, Multimethod Examination With Application to Gangs

“…Although few in number, these studies add great value to our understanding of the strengths and limits of using different data sources. Still, it is essential for scholars to continue to concern themselves with measurement because of its value for disciplinary advancement (Sullivan & McGloin, 2014). This call takes on added significance when examining substantive areas where little is known about measurement properties, such as gangs and gang members in prison.…”