This article reviews instruments that mental health clinicians commonly use to assess psychological symptoms, personality traits, behavioral problems, and psycholegal issues relevant to the assessment and management of criminal offenders. The instruments reviewed encompass a variety of formats, including self-report inventories, clinician rating scales, and structured clinical interviews that assess the presence of mental disorders, risk of violence and sexual offending, and response style.
Assessment of Psychiatric Symptoms and Mental Disorders
The high frequency of mental disorders among inmates in the United States is well established. Assessing the presence and nature of psychiatric symptoms often impacts offender management while incarcerated.
Self-Report Inventories
Multiscale self-report inventories have multiple advantages, including group administration, computerized scoring and interpretation, and broad coverage of symptoms and mental disorders. The Minnesota Multiphasic Personality Inventory-2 (MMPI-2), its revised, short-form (MMPI-2-RF), and the Personality Assessment Inventory (PAI) are frequently used in forensic settings. These instruments are intended for use with adults who have at least a fourth- to sixth-grade reading level and assess a range of symptoms, maladaptive personality traits, and behavioral problems. Although the MMPI-2 has more research support, the PAI may have broader application in correctional settings because it requires a lower reading level and offers correctional normative data that can inform judgments of risk of institutional misconduct and treatment needs. Versions of these instruments have also been developed for adolescents (i.e., MMPI-2-A, PAI-A) with item content modified to reflect developmental differences between adolescents and adults.
Screening Measures
Brief screening measures are used to identify those who require a more comprehensive psychiatric evaluation and may be useful when evaluating large numbers of individuals. Typically, screening measures are administered orally and require fewer than 5 min to complete. For example, the Brief Jail Mental Health Screen contains only eight questions assessing current mental health symptoms and treatment history. The Brief Jail Mental Health Screen was developed for use with adults and has demonstrated good predictive accuracy in identifying offenders with a mental disorder, although it may be more accurate for men than for women. A widely used screening tool for youth is the Massachusetts Youth Screening Instrument-Version 2. The Massachusetts Youth Screening Instrument-Version 2 has separate forms for boys and girls and is typically used upon entry to a juvenile justice facility to identify the mental health needs that can (or should) be addressed.
Cognitive Assessment Instruments
Cognitive disorders, including both intellectual disability (ID) and traumatic brain injury, are common in offender populations and can hinder the accuracy of other diagnostic approaches (e.g., self-report inventories). ID is characterized by significant deficits in intelligence and adaptive functioning (i.e., the extent to which an individual can function independently in activities of daily life). Although no measures have been developed specifically for assessment of cognitive functioning in correctional settings, a wide range of cognitive/ neuropsychological instruments can be used for this purpose. However, as these instruments are vulnerable to insufficient effort, cognitive assessment tools should be accompanied by measures of effort to ensure that the individual has performed to the best of his or her abilities. Similarly, no measures of adaptive functioning have been developed for correctional settings although a number of such scales exist, and the assessment of adaptive functioning is further complicated in correctional or forensic settings because offenders have little opportunity to demonstrate adaptive functioning abilities.
Risk Assessment Instruments
Another major focus of mental health assessment in the criminal justice system is violence and recidivism risk. Risk assessments influence both pre- (e.g., sentencing) and post-adjudication issues (e.g., parole eligibility). Research has repeatedly shown poor accuracy rates for risk assessments based on clinical judgment alone; the use of risk assessment instruments can help the clinician structure his or her decision-making, either through the application of a predetermined algorithm to gauge risk level (i.e., an actuarial approach) or identification of risk factors to be considered (i.e., the structured professional judgment [SPJ] approach). Violence risk assessment instruments also differ in the types of risk factors they include. Some include only static risk factors or those that relate to an offender’s history or personal characteristics that are stable over time, whereas others include dynamic risk factors or those that may change over time.
SPJ Instruments
The SPJ approach to risk assessment is best exemplified by the Historical-Clinical-Risk Management-20 (now in its third revision; HCR20v3). Intended for use with adults in correctional or mental health settings, the HCR-20V3 contains 20 static and dynamic risk factors for violence grouped into Historical (e.g., history of violence), Clinical (e.g., current mental health symptoms), and Risk Management (e.g., future problems with treatment adherence) categories. Clinicians rate the presence and relevance (for the individual being evaluated) for each risk factor and then use these ratings to develop a risk formulation, identify scenarios in which violence might occur, and generate risk management strategies. Clinicians also generate three summary risk ratings (Case Prioritization, Serious Physical Harm, and Imminent Violence—each rated as low, moderate, or high) to communicate the individual’s violence risk. A large body of research has demonstrated the utility of the HCR-20 with criminal offenders and forensic psychiatric patients. Other SPJ instruments target specific types of violence (e.g., sexual offending) or focus on a wide array of problem behaviors. For example, the Short-Term Assessment of Risk and Treatability contains 20 dynamic items that are used to evaluate risk of multiple problem behaviors, including violence, suicide, self-harm, recidivism, and treatment noncompliance.
Several SPJ instruments have been developed to facilitate the evaluation of risk in youth. For example, the Structured Assessment of Violence Risk in Youth comprises 24 static and dynamic risk factors and, unlike most SPJ instruments, includes six protective factors. Like the HCR-20, clinicians use these ratings to evaluate the youth’s overall risk level and identify potential targets for intervention. Similarly, an adolescent version of the Short-Term Assessment of Risk and Treatability is used to assess multiple domains of risk in juveniles.
Meta-analytic studies have consistently supported the predictive accuracy of SPJ tools. However, the reliance on individual case analysis and the need for specialized training often limit the applicability of SPJ approaches in correctional settings, where large numbers of offenders require a prerelease assessment.
Actuarial Risk Assessment Instruments (ARIAs)
ARIAs structure both the identification and weighting of risk factors and generate a probabilistic estimate of violence risk. Most ARIAs rely exclusively on static risk factors, although some include dynamic risk factors and/or assess change over time. One widely used ARIA is the Violence Risk Appraisal Guide (VRAG), and its revision, the VRAG-R. The VRAG was developed for use with adult offenders or forensic psychiatric patients and contains 12 static risk factors targeting demographic/personal background and health/ criminal history variables. Total scores on the VRAG/VRAG-R are grouped in one of nine bins that are associated with a probability of future violence. However, the VRAG, like most ARIAs, fails to account for change over time and omits idiosyncratic risk (or protective) factors that could impact an individual’s risk.
A widely used ARIA for gauging risk of recidivism is the Level of Service Inventory (LSI). The LSI is based on the Risk-Need-Responsivity model, which provides that risk management strategies are most effective when they are proportionate to the offender’s risk level, address the individual’s criminogenic needs, and account for characteristics that may impact how the offender responds to treatment. The LSI family of instruments includes a case management version (LS/CMI), a youth version (YLS/CMI), and a screening version. These tools comprise static and dynamic risk factors grouped into eight subcomponents (e.g., Criminal History, Education/Employment). These instruments generate a total score quantifying the offender’s overall risk/need level and eight subcomponent scores used to identify risk factors that may be targeted through treatment or supervision. Unlike many ARIAs, the LS/CMI and YLS/CMI encourage clinicians to develop their own estimate of risk based on the combination of their clinical judgment and actuarial data.
Psychopathy Instruments
As part of a risk assessment, clinicians often assess psychopathy, a personality construct characterized by interpersonal (e.g., dishonesty), affective (e.g., callousness), lifestyle (e.g., irresponsibility), and behavioral (e.g., criminality) traits. Research indicates that psychopathy is one of the strongest predictors of violence and general reoffending among adult offenders, although evidence supporting this construct is more mixed in juveniles. The most commonly used psychopathy instruments are those developed by Robert Hare and colleagues: the Psychopathy Checklist– Revised, Screening Version, and Youth Version. Individual Psychopathy Checklist Item scores are summed to yield a total score that reflects the severity of psychopathic traits. Self-report measures of psychopathy have also been developed, such as the Psychopathic Personality Inventory–Revised. However, given the risk of deliberate distortion, self-report measures are typically only utilized for research.
Sex Offender–Specific Instruments
A number of SPJ and ARAI tools have been designed for the assessment and management of sexual recidivism risk. Adult sex offender–specific SPJ tools include the Sexual Violence Risk-20 and the Risk of Sexual Violence Protocol. Modeled after the HCR-20, these tools guide clinicians through a series of ratings regarding the presence and relevance of static and dynamic risk factors, culminating in a summary rating of the offender’s overall sexual recidivism risk (low, moderate, or high). Although a modest literature has demonstrated strong support for these tools, the existing research is far less robust than for the HCR-20. The Static-99 (and expanded version, Static-2002) is the most commonly used sex offender–specific ARIA. The Static-99/Static-2002 (both of which were revised in 2009; Static-99R/Static-2002R) was developed for the assessment of sexual recidivism risk in adult males and assess static risk factors (with the exception of age). More recently, the Violence Risk Scale–Sexual Offender Version was developed to incorporate dynamic risk factors and assess change over time, while still maintaining the actuarial approach to decision-making. Scores generated by the Violence Risk Scale–Sexual Offender Version (static and dynamic Items) generate both probabilistic and qualitative (e.g., low, moderate–low) risk judgments.
A parallel set of risk assessment tools has been developed for juvenile sex offenders. Among the most widely used are the Juvenile Sex Offender Assessment Protocol–Revised and the Estimate of Risk of Adolescent Sexual Offense Recidivism.
The Juvenile Sex Offender Assessment Protocol– Revised contains 28 items that encompass both static and dynamic risk factors and form two static scales (Sexual Drive/Preoccupation; Impulsivity/Antisocial Behavior) and two dynamic scales (Intervention; Community Stability and Adjustment). The Estimate of Risk of Adolescent Sexual Offense Recidivism consists of 25 items divided into five domains: history of sexual assault, sexual interests and behaviors, psychosocial functioning, family/environment, and treatment. As with the HCR-20, clinicians use their item ratings on the
Juvenile Sex Offender Assessment Protocol– Revised and Estimate of Risk of Adolescent Sexual Offense Recidivism to formulate an overall risk estimate. Because there is considerable variability in estimates of their utility, these tools should be used as an adjunct to a thorough, individualized assessment of risk.
Response Style Instruments
An important component of any forensic mental health evaluation is the assessment of response style, including malingering/symptom exaggeration and defensiveness/symptom minimization. Numerous instruments have been designed to aid clinicians in the assessment of malingering and defensiveness, providing critical data about the validity of an offender’s self-reported symptoms or impairment. These instruments are typically divided into those that address cognitive functioning (i.e., cognitive effort) and those that assess validity of reported psychiatric symptoms.
Detecting Exaggerated Cognitive Impairment
Instruments designed to detect feigned or exaggerated cognitive impairment rely on a number of different approaches to detect suboptimal effort. Some measures use the floor effect, consisting of overly simple tasks that even those with severe cognitive impairment should complete. For example, the Rey 15-Item Memory Test requires the examinee to reproduce 15 letters, numbers, and shapes arranged in a manner that facilitates memory (e.g., A, B, C, 1, 2, 3). The Rey 15-Item Memory Test and other floor effect measures (e.g., the Dot Counting Test) have limited accuracy in identifying feigned cognitive impairment because examinees often recognize the simplicity of the task. Also, some types of genuine cognitive impairment and/or psychiatric symptoms can affect performance on these tests.
Another strategy used by cognitive effort tests is symptom validity testing (SVT), which compares the examinee’s performance on a forced-choice task to chance performance. The Test of Memory Malingering is a widely used SVT instrument that involves the presentation of 50 drawings. The examinee is subsequently provided with 50 recognition trials, each containing a previously seen drawing (target) and a distractor, and is asked to identify the target. Scores on the Test of Memory Malingering and other SVTs can be compared to chance performance (i.e., less than 50% correct) but can also be compared to normative data based on genuinely impaired individuals. Research has demonstrated strong classification accuracy for the Test of Memory Malingering in individuals with psychosis and traumatic brain injury but a somewhat greater risk of false positives (i.e., incorrect identification of feigning/ exaggeration) with adults with dementia or ID.
The Validity Indicator Profile integrates the floor effect and SVT approaches with the performance curve approach. Performance curves analyze an examinee’s performance on items of varying difficulty, with the expectation of increasing errors as item difficulty increases. The Validity Indicator Profile includes verbal and nonverbal subtests and classifies an examinee’s performance according to effort (high or low) and intention (to perform well or to perform poorly).
There are also cognitive effort tests derived from many commonly used tests. An example of an embedded cognitive effort test is the Reliable Digit Span index, a floor effect measure based on an individual’s performance on the Wechsler Adult Intelligence Scale, Fourth Edition (WAIS-IV) Digit Span subtest. Digit Span requires the examinee to remember progressively longer strings of numbers and recall them both forward and backward. The Reliable Digit Span is calculated by summing the longest series of digits recalled both forward and backward. Established cut scores generate good predictive accuracy but, like most cognitive effort measures, may misclassify individuals with ID.
Detecting Feigned Psychiatric Symptoms
Measures to detect feigned psychiatric symptoms include both interview-based and self-report scales and can also be embedded within existing symptom inventories or free-standing instruments. Clinicians commonly use the MMPI-2/RF and PAI to detect feigning. Importantly, these measures also permit clinicians to differentiate deliberate symptom exaggeration from random or confused responding and have been translated into numerous languages, making them useful in cross-cultural evaluations. However, the self-report nature of these instruments limits their use with offenders with severe cognitive impairment or psychosis, as they may not have the literacy, comprehension, or attention span necessary to complete these tests.
Commonly used free-standing measures of symptom exaggeration include the Structured Interview of Reported Symptoms-2 and the Miller Forensic Assessment of Symptoms Test. These instruments use a structured interview format and use multiple strategies to detect feigning (e.g., endorsement of rare symptoms, unusual symptom combinations, discrepancy between reported and observed symptoms). Whereas the Structured Interview of Reported Symptoms-2 utilizes a scoring algorithm to classify response style, the Miller Forensic Assessment of Symptoms Test, a screening tool, uses a total score to identify cases that require more in-depth assessment. Research has found that the Structured Interview of Reported Symptoms-2 has good predictive accuracy in identifying feigned psychosis but questionable utility in identifying feigned mood or anxiety disorders.
Detecting Minimization or Defensiveness
Clinicians frequently use the MMPI-2 and PAI to assess symptom minimization or defensiveness. Both instruments contain scales that assess denial of minor character flaws, exaggerated ability to cope with stress, and minimization of emotional distress. Clinicians also sometimes use free-standing measures of minimization or defensiveness, such as the Paulhus Deception Scales. The Paulhus Deception Scales is a 40-item self-report measure that assesses two types of defensive responding: conscious denial or defensiveness (Impression Management) and unconscious exaggeration of positive attributes (Self-Deceptive Enhancement). Research has found that MMPI-2, PAI, and Paulhus Deception Scales have good classification accuracy in identifying defensive responding.
Trial-Related Abilities
Competence to Stand Trial
The Supreme Court defined competence to stand trial in Dusky v. United States (1960), holding that a criminal defendant must have “sufficient present ability to consult with his attorney with a reasonable degree of rational understanding and a rational as well as factual understanding of the proceedings against him” (p. 402). A handful of instruments have been developed to assist clinicians in evaluating these abilities.
The MacArthur Competence Assessment Tool– Criminal Adjudication (MacCAT-CA) is a semistructured interview intended for use with adult offenders. Items on the MacCAT-CA are divided into three groups: (1) Understanding Items, which assess factual understanding; (2) Appreciation Items, which assess rational understanding; and (3) Reasoning Items, which assess ability to assist counsel. Understanding and Reasoning Items contain vignettes related to a hypothetical case, whereas Appreciation Items relate to the examinee’s own criminal case. The MacCAT-CA generates numerical scores for each subscale but does not generate an overall index of competence or a determination of competency.
Like the MacCAT-CA, the Evaluation of Competency to Stand Trial–Revised is designed to assist clinicians in assessing competence-related abilities in adult offenders. Evaluation of Competency to Stand Trial–Revised Items are divided into three scales that assess the three primary competence-related abilities and a fourth scale designed to assess response style (Atypical Presentation). Unlike the MacCAT-CA, Evaluation of Competency to Stand Trial–Revised items do not use vignettes but rather assess an offender’s legal understanding and ability to consult with counsel in his or her own case. Items also assess psychotic thought processes that may impact an offender’s rational understanding and ability to consult with counsel. Item scores are summed to generate three scale scores, and cut scores are used to classify the offender’s level of impairment.
Competence to Waive Miranda Rights
Another issue that arises in forensic evaluations where a defendant has made a confession is the capacity to understand Miranda rights. In Miranda v. Arizona (1966), the Supreme Court has ruled that a criminal defendant must understand his or her right to silence, utterances can be used against him or her, and the defendant has the right to an attorney. The most widely used scales to assess these rights were developed by Tom Grisso in the 1970s and recently revised. Like many forensic assessment instruments, Grisso’s Instruments for Assessing Understanding and Appreciation of Miranda Rights Scales are used to inform, but cannot definitively determine an individual’s capacity, and are vulnerable to deliberate distortion or exaggeration.
Criminal Responsibility
In many jurisdictions, the defense of not guilty by reason of insanity is available. This defense is defined differently across jurisdictions; but generally, a defendant raising it must establish that he or she did not understand the wrongfulness of or could not control his or her behavior because of a mental disorder. The Rogers Criminal Responsibility Assessment Scales was specifically designed to assist clinicians in these evaluations and comprises 25 scales that assess factors relevant to the not guilty by reason of insanity defense. Clinician ratings on these scales generate six summary ratings: potential malingering, presence of cognitive impairment, presence of a major mental disorder, loss of cognitive control, loss of behavioral control, and judgment of whether loss of cognitive/ behavioral control resulted from an organic condition or mental disorder. The psycholegal criteria ratings are then entered into a decision tree that generates a finding regarding the offender’s mental state at the time of offense (i.e., insane, sane, or no opinion). The Rogers Criminal Responsibility Assessment Scales has displayed good interrater reliability and concordance with not guilty by reason of insanity court verdicts in research; however, the research base supporting this instrument is very limited. Moreover, this tool has been criticized for failing to provide objective criteria to guide clinician ratings and is more useful in helping to organize an assessment of relevant information rather than providing a quantitative assessment of a defendant’s mental state.
Final Thoughts
Although numerous psychological tests are used with criminal offenders, the importance of the clinical issues involved in these evaluations requires reliable, well-validated instruments. In addition, consideration of cultural differences and educational background, which can impact the accuracy of these measures, is paramount.
References:
- Harris, G. T., Rice, M. E., Quinsey, V. L., & Cormier, C. T. (2015). Violent offenders: Appraising and managing risk (3rd ed.). Washington, DC: American Psychological Association.
- Heilbrun, K. (2001). Principles of forensic mental health assessment. New York, NY: Kluwer.
- Melton, G., Petrila, J., Poythress, N., & Slobogin, C. (2007). Psychological evaluations for the courts (3rd ed.). New York, NY: Guilford.
- Rogers, R. (2008). Clinical assessment of malingering and deception (3rd ed.). New York, NY: Guilford.
Court Cases:
- Dusky v. United States, 362 U.S. 402 (1960).
- Miranda v. Arizona, 384 U.S. 436 (1966).