Trajectories of Victimization From Early To Mid-Adolescence

• Date: 01 January 2011
• Author: Graham C. Ousey,Pamela Wilcox,Christopher J. Sullivan
• DOI: 10.1177/0093854810386542
• Published date: 01 January 2011
• Subject Matter: Articles

85

CRIMINAL JUSTICE AND BEHAVIOR, Vol. 38 No. 1, January 2011 85-104

DOI: 10.1177/0093854810386542

© 2011 International Association for Correctional and Forensic Psychology

AUTHORS’ NOTE: This research was sponsored, in part, by Grant DA-11317 (Richard R. Clayton, PI) from

the National Institute on Drug Abuse. The authors thank Richard R. Clayton, Scott A. Hunt, Michelle Campbell

Augustine, Shayne Jones, Kimberly Reeder, Staci Roberts Smith, and Jon Paul Bryan for their contributions to

the Rural Substance Abuse and Violence Project, which provides the data analyzed here. Direct all correspon-

dence to Christopher J. Sullivan, University of Cincinnati, School of Criminal Justice, 600 Dyer Hall, P.O. Box

210389, Cincinnati, OH 45221-0389; e-mail: christopher.sullivan@uc.edu.

TRAJECTORIES OF VICTIMIZATION

FROM EARLY TO MID-ADOLESCENCE

CHRISTOPHER J. SULLIVAN

PAMELA WILCOX

University of Cincinnati

GRAHAM C. OUSEY

College of William and Mary

A rapidly growing body of criminological research focuses on longitudinal trajectories of offending, with the aim of explor-

ing stability and change in antisocial behavior. A particularly intriguing debate within this area involves the issue of whether

there are multiple classes of offenders defined by distinct longitudinal patterns of offending. Parallel research on criminal

victimization, however, is lacking, with few studies exploring potential variation in individual trajectories of victimization.

The current analysis uses data from a panel of nearly 4,000 adolescents observed across a 4-year period to address this ques-

tion. The authors examined whether there are distinct classes of victimization trajectories across this time period. The analysis

revealed four groups. Descriptive analyses for key correlates of victimization were then conducted to explore their potential

correspondence with those of the observed victimization classes. The findings have implications for theory and empirical

research regarding between-individual differences and intraindividual change in victimization.

Keywords: victimization; latent classes; school crime

Research suggests that criminal victimization is a highly nonrandom, concentrated phe-

nomenon. Within the general population, most individuals experience no incidents of

victimization, whereas on the other end of the spectrum, a small group of persons experi-

ence victimization repeatedly, thereby accounting for a large percentage of the total number

of victimizations reported. Using repeated cross-sectional waves of the British Crime

Survey, for instance, Farrell and Pease (1993) demonstrated that approximately 60% of

persons surveyed reported no victimization experiences, whereas just more than 4% of

persons surveyed reported five or more victimization incidents in a single wave. This small

group of repeat victims accounted for approximately 45% of all reported victimization

incidents. Lauritsen and Davis Quinet (1995) revealed similar clustering in analysis of the

National Youth Survey, where 5% or less of sampled youth accounted for 39% to 63% of

the total reported victimization incidents within four different crime categories (assault,

larceny, vandalism, and robbery). More recently, analysis of data from a nationally repre-

sentative sample of youth showed that 62% were classified as nonvictims or “sparsely

victimized,” whereas 11% were victimized at relatively high rates across multiple types of

victimization (Reid & Sullivan, 2009). This clustering of high-frequency victimization

within a relatively small subset of individuals is remarkably similar to a well-established

86 CRIMINAL JUSTICE AND BEHAVIOR

pattern in studies of offending, in which a small percentage of individuals account for a

majority of all offenses reported in a sample (e.g., Piquero, Farrington, & Blumstein, 2007;

Shannon, 1980, 1988; Tracy, Wolfgang, & Figlio, 1990).

Despite this phenomenon of “hot victims” (or “hot dots”) within cross-sectional data, we

have little knowledge about the extent of intraindividual stability or change in victimization

patterns, particularly for individuals who differ in terms of the noted pattern of victimiza-

tion “concentration.” Although longitudinal research on victimization exists, it has tended

to model growth or change in victimization across sample respondents as a whole rather

than investigate the possibility of different patterns of change among subgroups of individu-

als. Furthermore, such longitudinal analysis has produced somewhat inconsistent evidence

in terms of how victimization at one time point correlates with victimization at a later time

point (cf. Lauritsen & Davis Quinet, 1995; Ousey, Wilcox, & Brummel, 2008). Although

these inconsistencies might be attributable to different samples or other methodological

variations across studies, there also is theory to support the idea that individuals may differ

in their temporal experiences with criminal victimization. For example, for a large share of

the population, initially low victimization levels may remain stable, whereas for a much

smaller group, initially high victimization levels may be retained across time. In contrast to

these images of relative stability, it also may be possible that some individuals with initially

low levels of victimization may experience sharp increases in victimization over time,

whereas others with initially high levels of victimization may see a temporal decline in their

victimization experiences because of changes in lifestyle or associates (e.g., Ousey et al.,

2008, for a recent review).

Unfortunately, to date, there have been few studies that explicitly investigate the possi-

bility of distinct intraindividual longitudinal trajectories of victimization. A rare exception

to this trend is found in a recent study by Higgins, Jennings, Tewksbury, and Gibson (2009),

which reports evidence of three distinct adolescent victimization trajectories. In addition,

they find that those trajectories appeared to be correlated with trajectories of an important

explanatory variable in the criminological literature, low self-control (Higgins et al., 2009).

Building on that analysis, the present study attempts to contribute to the limited empirical

research on this topic by using a latent class modeling approach to analyze victimization

trajectories across four waves of data from a panel of nearly 4,000 youth. Several specific

objectives are pursued. First, the analysis seeks to determine whether there are classes of

individuals who exhibit distinct longitudinal patterns of stability and/or change in victimiza-

tion. If so, to what extent do these classes resemble those reported in the innovative Higgins

et al. study? Furthermore, assuming that distinct victimization trajectories do emerge, a

second objective is to examine the extent to which the victimization trajectory of the dif-

ferent latent classes corresponds with trends in a dimension of low self-control as well as

several additional covariates of victimization, including self-reported delinquent behavior,

delinquent-friend exposure, illicit opportunities, and school bonds.

VICTIMIZATION THEORY AND GROUP TRAJECTORIES

Although most prior victimization research is cross-sectional, there are theoretical argu-

ments for expecting that criminal victimization may vary within individuals over time.

Moreover, there also is good reason to suspect that within-individual variation in victimization