Risk Assessment in Forensic Psychology

Risk Assessment in Forensic Psychology

• Risk assessment is an evaluation of a person's cognitive, emotional, and behavioral functioning.
• The goal is to obtain information that leads to a better understanding of the individual.
• This information is used to guide decisions, interventions, and management of offenders.
• Risk assessment considers the information used to predict risk.
• Risk prediction methods assess the likelihood someone will commit a crime in the future based on this information.
• Risk management creates effective strategies to manage the risk.

Assessment Measures

• Clinical interviews are used as assessment measures.
• Self-reports are used as assessment measures.
• Rating scales and checklists are used as assessment measures.
• Peer ratings are used as assessment measures.
• Direct observations are used as assessment measures.
• Actuarial instruments are used as assessment measures.
• Physical exams are used as assessment measures.
• Lab tests are used as assessment measures.
• Psychological instruments are used as assessment measures.

Why Do Assessments?

• Assessments are used in civil settings such as fitness to stand trial.
• Assessments are used in sentencing decisions.
• Examples of sentencing decisions include assessments of young offenders and dangerous offenders.
• Assessments are used in cases involving suicidal ideation.
• Assessments of psychological disorders are used.
• Assessments are used for criminal classifications.
• Assessments are used to create treatment intervention strategies.
• Assessments are used in parole decision-making.
• Assessments are used in pre-release decisions and planning.

Sources of Information

• Structured interviews are used to gather information.
• Self-reports are used to gather information.
• Collateral contacts from family and friends provide information.
• Police reports are a source of information.
• Prior criminal justice reports are sources of information.
• Court transcripts provide information about previous court cases.
• Judge's Reasons for Sentencing provide information.
• Pre-Sentence Reports provide information.
• Victim Impact Statements provide information.
• Prior mental health reports are sources of information.
• Psychological tests provide information.
• Actuarial measures provide information.

What is Being Assessed?

• Criminogenic risk factors (static) are factors that are not likely to change.
• Criminogenic needs (dynamic) are factors likely to change with suitable intervention.
• Stable vs. acute dynamic factors are considered as part of criminogenic needs.
• Treatment intervention strategies using identified needs as treatment targets can be important.
• Non-criminogenic needs may also be treated as part of general well-being.

Assessment Challenges

• Informed consent (voluntary, informed, and understood) is critical.
• Limits of confidentiality between researcher and clinician are considered.
• Rapport building is important.
• Earning trust is crucial in assessment.
• Evaluation of honesty is often important.
• Using collateral sources of information is a way to overcome these challenges.
• Standardized actuarial instruments are one way.
• Comparing information from various sources is an additional step.

Civil Setting: Duty to Warn/Protect

• Professionals can be liable if they fail to protect potential victims.
• The Canadian Code of Ethics requires psychologists to prevent serious physical harm or death of others, which may involve reporting to appropriate authorities or intended victims.

Approaches (Risk Prediction Models)

1. Unstructured clinical judgment

• First generation of risk prediction models.
• Information selection/combination is not guided by explicit rules.
• Decision rules are unclear.
• Often informal and subjective.
• Low accuracy.
• Problems include illusory correlations and reliance on base rates and salient cues.

2. Actuarial

• Second generation of risk prediction models.
• Pre-specified risk factors are entered into a statistical model.
• This combines and weights the factors.
• Formal, objective, empirically derived factors are used.
• Specific cutoffs are used for decisions.
• Focuses primarily on static measures which are not affected by intervention.
• Actuarial models tend to be more accurate than unstructured clinical judgment.

3. Structured Professional Judgment (SPJ)

• Third generation of risk prediction models.
• Uses specific risk factors derived from literature and includes static and dynamic aspects.
• Explicit rules guide selection of case-critical items but not item combination.
• Additional items may be considered that may not be part of the main risk measures.
• Rater makes a final decision about the risk level.

Specific Example Instruments (Risk Prediction Models)

SIR-R

• Used to assess risk of reoffending.
• Assesses various factors like the type of current offense, age at admission, previous incarcerations, security classification etc.

VRAG

• Designed to predict violent recidivism for serious offenders.
• Based on a large sample of forensic psychiatric patients.
• Contains 12 static items.
• Scores range from -27 to +35 using 9 risk bins, with higher scores indicating higher risk.

LSI-R

• Predicts general recidivism.
• Includes 54 risk and criminogenic needs items (static and dynamic).
• 10 subcomponents.
• Scores based on the risk/need levels are categorized into 5 levels, with higher scores indicating higher risk.

HCR-20

• Structured professional judgment instrument for violence risk assessment.
• Items are selected based on literature review and clinical experience.
• The model is based on the historical, present clinical, and future risk management items.
• Uses static and dynamic risk factors to predict risk level.

Outcome Statistics

• The prediction of risk is evaluated and statistically analyzed to generate data about correlation, odds ratio, analysis of variance, and survival curve analysis.

Decision Outcomes

• Assessing correct classifications are important.
• True Positive(violent individuals correctly identified).
• True Negative(non-violent individuals correctly identified).
• False Positive (non-violent individuals incorrectly identified as violent).
• False Negative (violent individuals incorrectly identified as non-violent).
• Various tools are used to analyze this data including ROC curves where AUC values are generated.

Risk Prediction Model Examples

• Baxstrom study examined the risk of mentally ill patients released into communities.

Communicating Risk Assessment Results

• Using risk levels to communicate results: Low, moderate, high, and extreme risk of future violence.

Treatment Principles

• Effective correctional treatment should be based on principles of risk, need, and responsivity (Andrews & Bonta, 2010).
• Assessing risk helps ensure interventions are focused on those at highest risk.
• Matching treatment to risk levels is an important principle to avoid misuse of resources.
• Identifying and treating criminogenic needs contributes to rehabilitation and preventing re-offending.
• Using interventions that match offender's learning styles helps improve effectiveness.

This quiz covers the key concepts of risk assessment in forensic psychology, including cognitive and behavioral evaluations. It discusses various assessment measures such as clinical interviews, self-reports, and rating scales. Understanding these elements is crucial for managing offenders and predicting criminal behavior.