Module 8 - Risk Assessments

For PSYO2444 (Forensic psychology)

Risk Factor

Measurable feature of an individual (e.g., attitude), predicts the behaviour of interest (e.g., recidivism) 

The components of risk assessment

Static risk factors and dynamic risk factors

Static risk factors (a component of risk assessment)

Past behaviour that cannot change (e.g., age at first arrest)

Dynamic risk factors (a component of risk assessment)

Current, changeable psychological features (e.g., antisocial attitude) 

The history of risk assessment

• Baxstrom v. Herald (1966): "Violent" offenders were released, with low recidivism rate.  

• Dixon v. Attorney General of the Commonwealth of Pennsylvania (1971): held that individuals found unfit to stand trial cannot be detained indefinitely and must either be restored to fitness, civilly committed, or released. 

• Baxstrom and Dixon studies: Base rate for violence is low but false positive rate is high. 

Important risk factors

Dispositional, historical, clinical, and contextual

Dispositional risk factors (an important risk factor)

Reflects traits, tendencies, or style. Demographics include age and gender while personality traits include impulsivity and psychopathy.

Historical risk factors (an important risk factor)

General history and specific criminal history, past antisocial behavior, age of onset, childhood history of maltreatment (were abused), past supervision failure (reoffended), escape (from custody), or institution maladjustment (hospital, jail, etc.) 

Clinical risk factors (an important risk factor)

Symptoms of mental disorders that can contribute to violence which includes substance use (support habits increasing the risk) and schizophrenia or affective disorders

Contextual risk factors (an important risk factor)

Aspects of the individual's current environment that can elevate risk which include lack of social support, access to weapons, and access to victims

Unstructured clinical judgement

Decisions are characterized by professional discretion and lack of guidelines. Does not look at specific risk factors and is based on clinicians' skills, experience, wisdom, and intuition. Is the most flexible approach with no rules regarding risk decisions and often involves a lengthy interview with the offender.

Limitations with Unstructured clinical judgements

Risk decision rules are unclear, subjective, includes illusory correlations, and clinicians are unable to predict violence at rates above chance (Manahan, 1981)

Actuarial Prediction/assessment

Decisions are based on risk factors selected and combined based on their empirical and statistical association. Can incorporate static and/or dynamic risk factors. Risk factors are weighted by importance for prediction and summed for total risk score), where they are often translated to % for risk to reoffend or categorized. Is more accurate compared to Unstructured clinical judgement.

Static-99R (actuarial prediction/assessment)

Most widely used sex offender risk assessment instrument developed by Hanson & Thornton. Includes 10 static risk items across age, living situation, index offence characteristics, and prior offence characteristics

Structured Professional Judgment (SPJ)

Bridges the gap between risk prediction and violence prevention. Instruments contain empirically-based static and dynamic risk factors with nominal response options (Absent, possibly present, present). It's accuracy is only starting to be assessed. Includes the Historical, Clinical, Risk (HCR-20).

Correct risk predicitons

True-negative (predicted to not reoffend, does not reoffend), and true-positive (predicted to reoffend, reoffends).

Incorrect risk predictions

False-positives (predicted to not reoffend, does not reoffend), and false-negative (predicted to reoffend, reoffends)

Reasons for incorrect risk predictions

False-positives affects the offender by wrongfully labelling them and denying them their freedom, and false-negatives affect society by letting a violent offender be free

Methodological issues in risk assessment

• Ideal evaluation vs. Reality: Assumes risk can be measured.  

• Weaknesses in research include the limited number of risk factors, how criterion variable is measured, and how the criterion variable is defined. 

Current issues in risk assessment

Protective factors, use of scientific research, why people stop reoffending, women offenders, and cross-cultural offenders

Protective factors (a current issue in risk assessment)

Factors that reduce or mitigate the likelihood of violence

Use of scientific research (a current issue in risk assessment)

Practitioners are not using instruments, and instead are relying on the MMPI which is a personality/psychopathology test (87%), while (<1%) are not using the LSI-R or VRAG (Level of Service Inventory-Revised, and Violence Risk Appraisal Guide): actual risk assessment tools used to predict the likelihood of reoffending

Why do people stop offending(desistance/while they get older)?(a current issue in risk assessment)

Large drop off in early adulthood, maturation (growing up), and situational factors

Women offenders (a current issue in risk assessment)

The same risk factors apply to women too, but it's too early to assume

Cross-cultural offenders (a current issue in risk assessment)

Ewert v. Canada (2015), where Ewert claimed Correctional Service Canada (CSC) use of assessments posed "clear danger" to Indigenous offenders who are "already disadvantaged."