Many of the forensic psychology students here at Carleton University are currently working on one or more projects, and have also contributed to research publications in the past. Below is a list of current and past thesis projects and publications.

Police Research Lab

Theses

Simon Baldwin

Title: Excited Delirium Syndrome (ExDS): Understanding the Issues and Reducing the Risks Associated with Police Use of Force

Abstract: The review of sudden and unexpected in-custody deaths (I-CDs) clearly demonstrates that there is a cluster of features which indicate that a subject is suffering from a medical emergency. For those who most frequently deal with these subjects (e.g., law enforcement, paramedics, emergency physicians, medical examiners) this is a real issue with serious implications. The labelling of this cluster of features as Excited Delirium Syndrome (ExDS) continues to be contentious. However a standardized and concise label with which meaning (e.g., medical emergency) can be assigned is necessary for the recognition, identification, intervention and treatment of these subjects. Additionally, despite there being many risk factors and a multitude of etiologies and pathophysiologies for ExDS, there are prevention and intervention strategies that can be employed within these dynamic and rapidly unfolding events to diminish adverse outcomes.

Through the theoretical lens of symbolic interactionism and the sociology of diagnosis, this research examines the meaning attributed to ExDS, how this meaning influences actions as well as the risks and benefits of medicalization. This research presents promising intervention strategies for gaining control of these subjects, as well as risk factors and officer safety concerns. Furthermore, through the use of grounded theory and excerpts from use of force reports, this research provides an interpretive account of the extreme and violent nature of encounters with probable cases of ExDS, providing a better understanding of these situations.

This research represents a new area of inquiry into non-fatal cases of ExDS and the prevention of sudden and unexpected I-CDs. The use of a mixed methods research design utilizes the strengths of qualitative and quantitative methods to analyze police use of force reports. This provides the opportunity to triangulate the results from each of these differing methodological approaches in order to elucidate, validate and generalize the findings. This gives both the depth and breadth required to inform law enforcement training and policy in the area of use of force and medically-high risk situations. As such, this research provides grounded recommendations for policy and training as a delivery mechanism of meaning as well as for equipment and use of force reporting. Thus, the overall focus and intent of this research is reducing the risk of I-CDs and improving police and public safety.

Brittany Blaskovits

Title: Walk This Way: A Kinematic Point-Light Investigation of Victim Vulnerability

Abstract: Recent research suggests that offenders may be able to detect vulnerability in others by observing gait patterns (e.g., Book, Costello, & Camilleri, 2013). It has been argued that offenders may use gait as a cue for victim selection. Study one seeks to explore the gait patterns of individuals who self-report a victimization history. The gait of male and female participants will be recorded using kinematic point-light display technology. To investigate potential explanations for the differential gait patterns of previous victims, the personality and affect of “walker individuals” will be examined. Bivariate correlation analyses and mediated regression are proposed to study the hypothesized relationships. In study two, a sample of observers will watch the point-light video recordings and rate the walkers on a number of characteristics including: vulnerability to future victimization, perceived submissiveness and dominance, and perceived positive and negative affect. A series of binomial logistic regressions will be used to predict the likelihood of observer accuracy in perceiving vulnerability, personality, and affect, from gait pattern. Determining the nonverbal cues that increase vulnerability is important for understanding the ‘hunting processes’ of offenders, as well as for reducing victim recidivism.

Holly Ellingwood

Title: How Should Across-Crime Similarity be Assessed in Behavioural Linkage Analysis? A Comparison of Jaccard, Sorensen-Dice and Simple Matching Across Two Different Crime Types and Sampling Methods

Abstract: The ability to link serial crimes accurately by relying on crime scene behaviours may be affected by the coefficient used to assess across-crime similarity. To examine this issue, the present study, the AUC was used to assess the level of linking accuracy that could be achieved when using one of three similarity coefficients – Jaccard’s coefficient (J), the simple matching index (S), and the Sorensen-Dice index (SD) – to measure across-crime similarity. Based on analyses using both serial burglary and serial rape data, the results demonstrated that, regardless of what measure of linking accuracy was used, no significant differences were found between the different similarity coefficients except in the case of S, in a single instance when it performed significantly worse than J and SD.   Secondly, the study sought to determine if sampling method (i.e., have an equal number of unlinked crimes to the linked crime or not limiting the unlinked crimes) impacted behavioural linkage accuracy. Both the r and the AUC were used to assess if the level of linking accuracy changed depending on sampling method used. Results showed that regardless of crime type of sampling method, the AUC levels were unchanged. However, r showed significant variability between sampling methods regardless of crime type, suggesting the type of linking measurement used may impact accuracy depending on the chosen sampling method.

Rebecca Mugford

Title: Using Classification Trees to Link Serial Crimes

Abstract: In the investigative setting, police must often decide whether multiple crimes have been committed by a single offender.  When forensic evidence is unavailable or compromised, crime analysts must depend on crime scene behaviours to make connections between crimes, a process commonly referred to as behavioural linkage analysis (BLA). Using a variety of statistical techniques (e.g., logistic regression analysis), a number of studies over the last few decades have shown that it is possible to link serial crimes in a relatively accurate fashion. However, despite the demonstrated value of these techniques, linking decisions in the investigative setting are still typically made in an unsystematic (and arguably inferior) way by relying on linking cues deemed relevant by a given analyst. The goal of the current study is to present a novel approach to BLA which relies on classification trees (CTs). Using two different types of crime data, this study will assess whether CT-based linking models (standard CT, ICT, and multiple ICT) can result in levels of predictive accuracy that are comparable to (or surpass) that achieved by logistic regression linking models (the currently preferred method for carrying out BLA). Not only is the CT approach predicted to link crimes with the same degree of accuracy as other statistical approaches, it is less complex and cumbersome than other approaches and more intuitively appealing, both in terms of how it can be used in practice by analysts and in terms of the linking strategies that it produces (e.g., strategies that are more idiographic in nature). Equally valuable, a CT approach may uncover important patterns of serial offending behaviour that have previously gone unnoticed using the traditional approaches to BLA because these approaches have relied on a main-effects, one-size-fits-all approach to examining behavioural consistency and distinctiveness.

The Aggressive Cognitions and Behaviour Research Lab

Theses

Sacha Maimone, M.A.

Title: The Relationship Between Attitudes Towards Violence and Violent Behaviour: The Use of Implicit and Self-Report Measures

Abstract: The current studies investigated the relationship between attitudes towards violence and violent behaviour.  Violent attitudes have mostly been assessed with self-report measures.  Within social psychology, implicit attitudes have also been assessed using response latency measures finding significance regarding these attitudes.  The current studies examined implicit and self-report attitudes, as well as the relationship between attitudes and past/future violence, among three studies (one containing offenders).  The effect of observing, or engaging in, violence on attitudes and whether this affects the relationship with violent behaviour was also examined.  No significant results were found involving implicit attitudes; however self-report attitudes were positively related to measures of violent behaviour and more positive self-report attitudes were found after observing, or engaging in, the violent task, as was a positive relationship between these attitudes and future violence.  These results extend previous research and provide valuable information regarding the role of attitudes in the commission of violence.

Cathrine Pettersen, M.A. 

Researcher: Factor Structure of the Aggression Questionnaire with Sexual Offenders

Abstract: Confirmatory factor analysis was used to examine the factor structure of the Aggression Questionnaire (Buss & Perry. 1992) in a sample of incarcerated adult men with a history of sexual offending (N =293). The main aim of the study was to determine whether factor models reported in previous studies for forensic- and non-forensic samples replicate with sexual offenders. Results of the current study did not lend support for the most commonly reported four- factor model of the full or reduced (12-item) versions of the Aggression Questionnaire. Alternative single- and two-factor models also did not provide adequate fit. From an empirical standpoint, results of post-hoc exploratory analyses suggested that a three-factor model within which the Verbal Aggression and Anger items were combined to form a single factor provided the best fit for the current data. Possible explanations for the findings and implications for theoretical conceptualizations of aggressiveness in sexual offenders are discussed.

Publications

Pettersen, C., Nunes, K. L., Woods, M., Maimone, S., Hermann, C. A., Looman, J., & Spape, J. (2014). Does change in hostility predict sexual recidivism? International Journal of Offender Therapy and Comparative Criminology. Advance online publication. doi: 10.1177/0306624X14547033

Nunes, K. L., Hermann, C. A., Maimone, S., & Woods, M. (2014). Thinking clearly about violent cognitions: Attitudes may be distinct from other cognitions. Journal of Interpersonal Violence. Advance online publication. doi: 10.1177/0886260514540329