Maryland Population Research Center

Assistant Professor
Criminology and Criminal Justice
University of Maryland
College Park, MD 20742

Email: hpetras@crim.umd.edu
Phone: 301-405-4699
Office: 2220 LeFrak Hall

CV

Publications

Grants

Recent Scientific Accomplishments

Over the past three years, Petras focused on three areas of research: (1) Nature and course of aggressive behavior and its related consequences in urban youth; (2) Quality and classification accuracy of screening tools for ` violence and other criminal acts; and (3) Application of latent variable methods to cross-sectional and longitudinal data sets. First, Petras has a series of publication relating the developmental trajectories of aggressive behavior to negative outcomes in young adulthood and how this relationship varies as a function of gender. For example, in Petras, et al. (2004), he was interested in understanding which predictors could be used for boys who have outcomes that are inconsistent with their early aggressive behavior growth trajectory. First grade reading achievement, race, and poverty status proved to be significant early predictors of discordance between early patterns of aggression and later outcomes; whereas the significant middle school predictors of discordance were parent monitoring, deviant peer affiliation, and neighborhood level of deviant behavior. Together with colleague Dr. Schaeffer, Petras further compared growth trajectories for girls and boys via multiple group growth mixture analysis (Schaeffer, Petras, Ialongo, et al, 2006) and found two analogous aggression trajectories for girls and boys: chronic high aggression and stable low aggression. Findings from these studies were published in top ranking journals including Developmental Psychology, Development and Psychopathology, Journal of Consulting and Clinical Psychology. Second, given the strong evidence of aggressive behavior as a precursor for later violence and other antisocial outcomes, prevention scientists have used measures of early aggressive behavior to identify children for inclusion in indicated violence preventive interventions. Building upon his training in prevention, Petras argued that the risk factor approach (i.e., using aggression as an antecedent of later violence) and the screening approach (i.e., using aggression to identify individuals at risk for later violence) are two closely related strategies, which can inform each other. Using data from the Baltimore Prevention Program, he investigated this issue in two publications (Petras et al., 2004; Petras, et al, 2005). Importantly, extending the traditional ROC approach, he incorporated the cost of false positives/false negatives into the decision-making about optimal cut points of the predictor variables (Kraemer et al., 1999, 2001). When comparing the results by gender, he found a number of similarities and differences. For both boys and girls, teacher ratings of aggression were significantly associated with the later juvenile court record. Importantly, in contrast to boys, screening for girls tends to become optimal only towards the end of elementary school. The exception is for the cheap and noninvasive intervention scenario, where for both boys and girls the screening can be done starting in first grade. This work was published in American Academy of Child and Adolescent Psychiatry. Third, Petras?interest in latent variable models has formed the basis for his empirical work.

Funded Research

Petras has a pending grant proposal (R01) to NICHD, with Child Trends, to examine trajectories of health risk behaviors in adolescence using the NLSY97 data set.

Future Research Plans

Petras will continue working on his research interests, but since joining the Department of Criminology will focus more explicitly on topics closely related to criminal behavior. For example, together with his student he applied Latent Class Analysis to victimization experience in order to determine if subgroups of victimized individuals existed. In addition, he has a project revisiting the Criminal Career concept to study variation in the level of versatility of criminal engagement. He is currently preparing a grant application investigating the co-occurring development of academic achievement and aggressive behavior in relation to later negative outcomes, such as school dropout, incarceration, and employment. He plans to lend his skills in developmental psychology to expand the focus of the Crime and Population Work Group.

Maryland Population Research Center
0124N Cole Student Activities Building (#162)
College Park, MD 20742
Phone: 301-405-6403
Fax: 301-405-5743