Skip to content
Search resources Open menu

Topics of Intimate Partner Violence

The HCR-20 V3 in Family Violence Cases

Published

Snap Shot

  • The question: What can the HCR-20 V3, a general violence risk instrument, tell a court in a partner or family violence case?
  • What the research shows: Studies report good to excellent interrater reliability for most ratings (for example, excellent for summary risk ratings in a German study of five raters and 30 cases). In prospective studies of discharged forensic patients and of 114 incarcerated men, accuracy for predicting general violence was moderate, with AUCs of .70 to .77. Broader meta-analyses place the HCR-20 among the better-performing tools but find that most instruments identify people unlikely to be violent better than they confirm that a high-risk person will be.
  • What it does not show: Validation samples are mostly forensic patients and prisoners, and outcomes are usually any violence. Published evidence on predicting partner violence is sparse, and no published study located for this paper establishes accuracy for violence toward children. Its low, moderate, or high summary ratings are qualitative, not probabilities, and the instrument does not show whether a charged act occurred.
  • The law: Admissibility turns on fit under Rule 702 (Daubert, Kumho Tire, Houser). At the guilt phase, risk evidence raises profile and Rule 404 concerns (Ketchner, Haskie); the instrument is used chiefly in release, sentencing, and commitment, and in custody matters only with care.
  • Issues for each side: Prosecutors should not treat a general rating as settling partner or child risk and should ask whether a partner-violence-specific tool was also used; defense counsel should not treat the instrument as lacking scientific support, and should focus on population fit and on ratings converted into percentages.

The HCR-20 Version 3 (HCR-20 V3) is a set of professional guidelines for assessing and managing the risk of violence. Published in 2013, it is a widely used and researched violence risk instrument in forensic psychiatry, corrections, and civil commitment. It was not designed for intimate partner or family violence, yet it appears in family violence cases: in evaluations of defendants with significant mental illness, in release and sentencing proceedings, in civil commitment and conditional release matters, and occasionally in custody disputes where a parent’s violence history is at issue.

Attorneys who encounter an HCR-20 V3 report need to understand what the instrument does, how its ratings and conclusions are reached, what the validation research shows, and why a general violence tool may miss features specific to partner violence. This paper describes the instrument’s structure and administration, summarizes research on its reliability and predictive validity through 2024, explains its limits in intimate partner and family violence cases, discusses when evaluators pair it with partner-violence-specific instruments, and reviews the legal framework for admitting and challenging such evidence as of November 2026.

What the HCR-20 V3 Is

The HCR-20 was first published in 1995 and revised in 1997 (Version 2); by the time Version 3 appeared, Version 2 had been the subject of more than 200 disseminations based on more than 33,000 cases in 25 countries (Douglas et al., 2014). The name reflects its three groups of risk factors: Historical, Clinical, and Risk Management. Version 3 kept the 20 core factors but revised their content, added sub-items and relevance ratings, and added steps for formulation and scenario planning (Douglas et al., 2013).

The HCR-20 V3 is a structured professional judgment (SPJ) instrument. In SPJ, the evaluator systematically considers a fixed list of factors identified by research and professional consensus, but does not add up a score to produce a probability. Instead, the evaluator integrates the information into a judgment about the level of concern and, most importantly, a plan for managing the risk. This contrasts with actuarial instruments, which combine items by a fixed formula and link total scores to reoffense rates observed in a development sample. The SPJ approach was designed to retain the structure and transparency that improve on unaided clinical opinion while allowing the evaluator to account for case-specific factors and to plan interventions (Hart et al., 2016).

The Risk Factors

The Historical scale contains ten factors describing a person’s past. Most HCR-20 V3 factors are framed as “problems with” a domain, and published case formats commonly use the item titles shown here:

  • H1. History of Problems with Violence
  • H2. History of Problems with Other Antisocial Behavior
  • H3. History of Problems with Relationships
  • H4. History of Problems with Employment
  • H5. History of Problems with Substance Use
  • H6. History of Problems with Major Mental Disorder
  • H7. History of Problems with Personality Disorder
  • H8. History of Problems with Traumatic Experiences
  • H9. History of Problems with Violent Attitudes
  • H10. History of Problems with Treatment or Supervision Response

The Clinical scale contains five factors describing recent functioning, typically the past several months:

  • C1. Recent Problems with Insight
  • C2. Recent Problems with Violent Ideation or Intent
  • C3. Recent Problems with Symptoms of Major Mental Disorder
  • C4. Recent Problems with Instability
  • C5. Recent Problems with Treatment or Supervision Response

The Risk Management scale contains five factors anticipating the future, rated for the specific context the person is expected to be in, such as remaining in an institution or living in the community:

  • R1. Future Problems with Professional Services and Plans
  • R2. Future Problems with Living Situation
  • R3. Future Problems with Personal Support
  • R4. Future Problems with Treatment or Supervision Response
  • R5. Future Problems with Stress or Coping

Each scale also allows the evaluator to note “other considerations,” meaning case-specific factors not captured by the listed items. A frequent source of confusion in reports is the treatment-response domain, which appears three times: H10 addresses the person’s history, C5 recent response, and R4 anticipated response. These are distinct ratings across time frames, not duplicates. The test publisher’s materials, not this paper, define what each item covers and how it is rated.

How Ratings Are Made

Each factor is rated for presence on a three-level scale: not present, possibly or partially present, or present. Evaluators may omit a factor when there is insufficient information. Each factor is then rated for relevance, meaning how important it is to this person’s violence, on a scale of low, moderate, or high. A factor can be present but of low relevance (for example, a remote employment history unrelated to the person’s violence) or highly relevant despite partial presence.

Presence and relevance ratings do not convert to a numerical score in clinical use. Researchers often sum presence ratings to create a total score for study purposes, which is why published validation studies report total-score results; but the intended conclusion is the evaluator’s summary judgment, not a sum.

Formulation, Scenarios, and Summary Judgments

Version 3’s most significant change is its emphasis on formulation: an explanation of why this person has been violent, what drives and sustains that violence, and what could reduce it (Hart & Logan, 2011). Formulation organizes the relevant factors into a coherent account, for example that a person’s violence toward partners has occurred during periods of heavy drinking and perceived abandonment, typically after separation, when psychiatric treatment had stopped.

The evaluator then develops scenarios, which are plausible descriptions of future violence. Commonly used scenario types include a repeat scenario (violence like past violence), an escalation or worst-case scenario (more serious harm), a twist scenario (a change in victim, motive, or method), and a best-case or improvement scenario. Each scenario specifies the likely nature, severity, victims, timing, and warning signs of violence, and the evaluator then recommends management strategies such as monitoring, treatment, supervision, and victim safety planning tied to those scenarios.

Finally, the evaluator makes summary judgments, recorded on the rating sheet as low, moderate, or high, for case prioritization (how much intervention is needed), risk of serious physical harm, and imminence of violence, along with a recommended reassessment date. These judgments are qualitative. They do not mean that a stated percentage of similar people will be violent.

Reliability and Validity Research

Reliability

Interrater reliability concerns whether two trained evaluators reviewing the same case reach the same ratings. In a study of 35 forensic psychiatric patients, rated by experienced clinicians, reliability for presence ratings and summary risk ratings was excellent, while most relevance and sub-item ratings were in the good-to-excellent range with a minority in the fair or poor range (Douglas & Belfrage, 2014). A German study in which five newly trained raters, none with prior SPJ experience, each rated the same 30 cases found excellent agreement on summary risk ratings and good agreement on item ratings, with some lower-reliability ratings (Kötter et al., 2014). A retrospective file study of 124 insanity acquittees (100 men and 24 women) found higher interrater reliability for men than for women, and associations between scale scores and violence that were higher among men, although sex did not significantly moderate the prediction of violence; the authors called for research on how scoring guidance is applied across sexes (Green et al., 2016).

Predictive Validity

Several prospective studies support Version 3’s predictive validity. In a study of 106 offenders and civil psychiatric patients leaving institutions, presence ratings, relevance ratings, and summary risk ratings were all associated with violence at follow-up, summary ratings added to the prediction from presence ratings, and results did not differ by gender or sub-sample; Version 3 correlated highly with Version 2 and predicted violence comparably (Strub et al., 2014). In England and Wales, Version 3 predicted community violence after discharge from medium-secure hospitals with AUCs of .73 at six months and .70 at twelve months (Doyle et al., 2014). The AUC, or area under the curve, is the probability that a randomly chosen person who was violent was rated higher than a randomly chosen person who was not; .50 is chance and 1.00 is perfect. Values around .70 are conventionally described as moderate accuracy.

Newer research has extended the findings across settings. A prospective study of 114 incarcerated men in Mexico City found that summary risk ratings and total scores predicted institutional violence over three months, with AUCs of .71 to .77 (Nijdam-Jones et al., 2021). In a German retrospective study of 452 women with substance use disorders discharged from forensic psychiatric care, the HCR-20 V3 showed good predictive accuracy for both inpatient violence and violent recidivism (Wolf et al., 2024). A meta-analysis of the HCR-20 in adult women, which included mostly Version 2 studies, found moderate associations with violence for the total score and recommended further research on Version 3 (Rossdale et al., 2020). Broader meta-analyses of violence risk instruments generally place the HCR-20 among the better-performing tools but find that most instruments identify people unlikely to be violent much better than they confirm that a high-risk person will be (Fazel et al., 2012; Singh et al., 2011).

Three features of this literature deserve emphasis. First, most validation samples are forensic psychiatric patients, prisoners, or civil psychiatric patients, not community partner-violence defendants. Second, outcomes are usually any violence, often inpatient or institutional, rather than violence toward an intimate partner or child. Third, much of the research has been conducted by the instrument’s authors or their collaborators; a systematic review found that studies authored by tool developers tended to report more favorable results than independent studies (Singh et al., 2013).

Limits in Intimate Partner and Family Violence Cases

The HCR-20 V3 assesses general violence risk, and partner violence has features the instrument does not specifically address. These include coercive and controlling behavior short of physical assault, stalking, threats directed at a partner, violation of protective orders, the escalating significance of separation, sexual violence within the relationship, and the victim’s circumstances, such as barriers to leaving or to safety planning. Some of these may be captured under H1, H3, H9, C2, or “other considerations,” but the instrument does not prompt for them systematically.

The HCR-20 V3 also gives considerable weight to mental disorder (H6, H7, C3) and to insight and treatment response. That emphasis suits forensic psychiatric populations but can be less informative for the many people charged with partner violence who have no major mental disorder. Research on partner violence by men has long distinguished those whose violence is confined to the family from those who are generally violent and antisocial (Holtzworth-Munroe & Stuart, 1994); that typology was developed from studies of men, and it has been less studied for women and for partners of the same sex. A general violence instrument could in principle assign lower concern to a family-only person with no criminal record outside the home and no psychiatric history, even if the partner-specific risk factors are serious; the studies located for this paper did not test whether it does.

The evidence comparing general and partner-violence-specific instruments is mixed. A meta-analysis of 18 studies of spousal assault recidivism found that general violence and recidivism instruments (such as the VRAG, PCL-R, and LSI-R) predicted about as well as spousal-assault instruments: the average effect size was d = .54 across four studies of general instruments and d = .40 across ten studies of spousal-assault instruments, with overlapping confidence intervals; that review did not report separate results for the HCR-20 (Hanson et al., 2007). Direct published evidence on the HCR-20 V3’s accuracy for predicting partner violence specifically remains sparse, and no published study located for this paper has established its accuracy for predicting violence toward children within the family. An evaluator relying on the HCR-20 V3 alone to forecast partner or child violence is therefore extrapolating from general violence research.

Pairing the HCR-20 V3 with Partner-Violence-Specific Instruments

For these reasons, evaluators commonly use the HCR-20 V3 alongside, not instead of, an instrument developed for partner violence when partner violence is at issue. The Spousal Assault Risk Assessment Guide (SARA), now in its third version, is an SPJ instrument designed for partner violence and includes victim vulnerability factors; its structure parallels the HCR-20 V3, which makes combined use straightforward. The Ontario Domestic Assault Risk Assessment (ODARA) is a brief actuarial instrument developed for men with a police record of assault on a female partner. The two approaches answer different questions: the ODARA ranks a man’s likelihood of police-recorded reassault relative to others, while SPJ instruments organize case formulation and management. Neither was developed for violence toward children, and neither is a test of whether a charged act occurred.

Pairing is most useful in cases involving significant mental illness, where the HCR-20 V3 helps explain how symptoms, insight, and treatment adherence relate to violence, and in generally violent people whose partner violence is part of a broader pattern. In cases where the violence is confined to the relationship and mental disorder is not prominent, a partner-violence-specific instrument will usually be the primary tool. Within the HCR-20 V3 itself, evaluators can tailor scenarios to partner and family violence and consider partner-violence-related attitudes under H9 and relationship history under H3. (For these instruments, see IPV Risk Assessment Instruments.)

Limits of the Evidence

Most HCR-20 V3 research reports group-level accuracy in samples unlike many family-violence cases; many studies are retrospective file reviews, several have small samples (35 to 124 participants in studies cited here), and sex-specific findings rest on small numbers of women. Reliability figures depend on rater training and the quality of records, and an AUC describes how well a rating ranks people across a group, not the chance that a particular person will be violent.

Questions to Ask About This Research

When AUCs of .70 to .77 are cited for the HCR-20 V3 (Doyle et al., 2014; Nijdam-Jones et al., 2021), ask what was predicted and in whom: those studies followed discharged forensic patients and 114 incarcerated men for any violence, not partner or child violence, and an AUC describes ranking across a group, not the probability for one person. When reliability is described as “excellent,” ask which ratings: in a study of 35 patients, presence and summary ratings were excellent but a minority of relevance and sub-item ratings were fair or poor (Douglas & Belfrage, 2014), and in a study of five raters and 30 cases, item-level agreement was good rather than excellent (Kötter et al., 2014). When the HCR-20 is said to be among the better-performing instruments, ask what that comparison measured, since the meta-analyses find better identification of low-risk than high-risk people (Fazel et al., 2012) and developer-authored studies tended to report more favorable results (Singh et al., 2013). When the instrument is used in a partner-violence or custody case, ask whether any study cited examined partner or child violence, and whether the evaluator also used an instrument developed for that purpose. These questions are explained in “Reading Research Claims: A Guide for Attorneys” on this site.

The HCR-20 V3 and the Law

Admissibility and Weight

Expert testimony relying on the HCR-20 V3 is governed by Federal Rule of Evidence 702 and Military Rule of Evidence 702, interpreted under Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993) and Kumho Tire Co. v. Carmichael (1999), which applies the gatekeeping obligation to clinical and experience-based expertise as well as laboratory science. The 2023 amendment to Rule 702 emphasizes that the proponent must show by a preponderance that the opinion rests on sufficient facts, reliable methods, and a reliable application of those methods. Arizona adopted the federal standard in its 2012 amendment of Arizona Rule of Evidence 702 and amended the rule again effective January 1, 2024 (Order R-23-0004) to conform to the 2023 federal amendment. In the military, Daubert applies through Military Rule of Evidence 702, and courts-martial also apply the factors identified in United States v. Houser (C.M.A. 1993) for admitting expert testimony, including the qualifications of the expert, the subject matter, the basis for the opinion, legal relevance, reliability, and the balance of probative value against unfair prejudice.

Because the HCR-20 V3 is widely used and has been extensively studied, the more substantial questions concern fit rather than whether it has any scientific basis: whether it was validated for the population (a community partner-violence defendant rather than a forensic patient), whether it predicts the outcome at issue (partner or child violence rather than any violence), whether the evaluator was trained and had adequate records and interview access, and whether the conclusions go beyond the qualitative summary judgments the instrument supports. Because the summary ratings are not probabilities, testimony converting “high” into a percentage likelihood of future violence would require a separate basis.

In Barefoot v. Estelle (1983), the Supreme Court declined to hold that predictions of future dangerousness are inherently too unreliable to be admitted, leaving their weight to cross-examination and opposing evidence. Structured instruments developed in part as a response to concerns about unaided clinical prediction. Barefoot predates Daubert and the current text of Rule 702, which requires the proponent to show that the method is reliable and was reliably applied.

Guilt-Phase Limits and Profile Concerns

Risk instruments describe characteristics associated with future violence; they do not show whether a charged act occurred. Arizona courts have warned against expert testimony that invites jurors to infer guilt from a defendant’s resemblance to an offender profile. In State v. Ketchner (2014), the Arizona Supreme Court found error where an expert’s testimony about how abusers react to a loss of control was tied to the defendant’s characteristics, not to any victim behavior at issue, and so invited a finding that the defendant matched a profile. In State v. Haskie (2017), the court held that “cold” expert testimony explaining victim behavior is not categorically inadmissible and found no error on that record, but directed trial courts to screen such testimony under Rules 401, 402, 403, and 702, to give a limiting instruction on request, and to keep experts from speaking in broad, categorical terms without empirical support. An HCR-20 V3 report offered at the guilt phase would raise similar concerns, as well as character-evidence issues under Rule 404, because many of its factors (prior violence, antisocial behavior, personality disorder) are the kinds of propensity evidence that Rule 404(b) restricts.

Release, Sentencing, Commitment, and Custody

The HCR-20 V3 appears most appropriately in proceedings where future risk is the question. Arizona’s release statute directs the judicial officer, in setting the method of release or the amount of bail, to consider evidence that the accused poses a danger to others in the community and, in a domestic violence charge, the results of any risk or lethality assessment presented to the court (A.R.S. § 13-3967(B)(4), (5)). In Simpson v. Miller (2017), the Arizona Supreme Court held that the constitutional and statutory provisions categorically denying bail to everyone charged with sexual conduct with a minor under fifteen were facially unconstitutional because the offense is not inherently predictive of dangerousness, so detention requires a case-specific inquiry; structured risk evidence can contribute to that individualized inquiry. At sentencing, the instrument’s management recommendations may inform supervision and treatment conditions. In civil commitment and conditional-release proceedings for persons with mental illness, the setting most similar to its validation samples, it is often central.

In family court, Arizona law requires the court to consider domestic violence in deciding legal decision-making and parenting time. As amended in 2026 (H.B. 2995, 2026 Ariz. Sess. Laws ch. 263), A.R.S. § 25-403.03 creates a mandatory rebuttable presumption that awarding legal decision-making or parenting time to a parent found to have committed domestic violence is contrary to the child’s best interests, and its definition of domestic violence now includes coercive control; counsel handling matters pending around the time of the amendment should confirm which version applies. An HCR-20 V3 evaluation may be offered on whether a parent poses an ongoing risk, but because the instrument has not been validated for predicting violence toward children or for custody decisions, its use there requires particular care about what it can and cannot show.

Special Issues in Legal Practice

The principles described above apply with equal force to every party in a legal proceeding. The science operates the same way regardless of who is asking the questions, and the research summarized here does not by itself favor either side. What differs between the parties is not the science but the vantage point — the particular places where these issues tend to surface, and the particular errors each party is best positioned to guard against. The two summaries below identify those recurring issues. They are offered for orientation, not as litigation strategy; in every case the controlling question is what the evidence in that case will, and will not, support, and the expert’s role is to inform the trier of fact rather than to advocate for either side.

Special Issues for Prosecutors

Prosecutors may encounter the HCR-20 V3 in release hearings, sentencing, and proceedings involving defendants with mental illness, sometimes in a report that rates a defendant at low or moderate concern. The research supports describing the instrument as well studied, with good-to-excellent interrater reliability for most ratings and moderate accuracy for predicting general violence in forensic psychiatric and correctional samples, and supports asking whether the evaluator was trained, what records and interviews informed the ratings, and whether the summary judgments follow from the factors identified.

The research does not support treating a general violence rating as settling the question of partner or child risk, or ignoring the absence of partner-violence-specific factors such as stalking, protective-order violations, coercive control, and recent separation. It is useful to ask whether an instrument developed for partner violence was also used, whether the scenarios addressed partner and family violence specifically, and whether the evaluator had access to the complainant’s account and the full incident history. Summary ratings should be presented as the instrument defines them, as qualitative judgments about priority and management, not as probabilities, and risk evidence offered at the guilt phase raises profile and character-evidence concerns.

Special Issues for Defense

Defense counsel may encounter the HCR-20 V3 in release, sentencing, commitment, and custody proceedings, often in a report that rates a client at high concern. The research supports asking whether the evaluator was trained, what records and interviews informed the ratings, whether relevance ratings and formulation were completed, and whether the summary judgments follow from the factors identified. It also supports attention to the dynamic and management factors (the C and R scales), which the instrument is designed to track over time and which may support conditions short of the most restrictive option.

The research does not support treating the instrument as lacking scientific support, since its reliability and general predictive validity are documented in the studies reviewed. It is useful to ask whether the validation research matches the case: most studies involve forensic patients and prisoners, and outcomes are usually any violence rather than partner or child violence. Questions about population fit, about treating qualitative ratings as percentages, and about the weight given to historical factors that cannot change are grounded in the literature; so are the limits of that literature for the defense, since a moderate AUC for a group does not show that a particular rating is wrong. Neither side should present the instrument as evidence of credibility or of whether a charged act occurred.

This paper is for educational purposes only and is not legal advice. The author is a psychologist, not an attorney; summaries of law are general and must be verified against current authority or with a licensed attorney.

References

Douglas, K. S., & Belfrage, H. (2014). Interrater reliability and concurrent validity of the HCR-20 Version 3. International Journal of Forensic Mental Health, 13(2), 130–139. https://doi.org/10.1080/14999013.2014.908429

Douglas, K. S., Hart, S. D., Webster, C. D., & Belfrage, H. (2013). HCR-20 V3: Assessing risk for violence: User guide. Mental Health, Law, and Policy Institute, Simon Fraser University.

Douglas, K. S., Hart, S. D., Webster, C. D., Belfrage, H., Guy, L. S., & Wilson, C. M. (2014). Historical-Clinical-Risk Management-20, Version 3 (HCR-20 V3): Development and overview. International Journal of Forensic Mental Health, 13(2), 93–108. https://doi.org/10.1080/14999013.2014.906519

Doyle, M., Power, L. A., Coid, J., Kallis, C., Ullrich, S., & Shaw, J. (2014). Predicting post-discharge community violence in England and Wales using the HCR-20 V3. International Journal of Forensic Mental Health, 13(2), 140–147. https://doi.org/10.1080/14999013.2014.906517

Fazel, S., Singh, J. P., Doll, H., & Grann, M. (2012). Use of risk assessment instruments to predict violence and antisocial behaviour in 73 samples involving 24 827 people: Systematic review and meta-analysis. BMJ, 345, e4692. https://doi.org/10.1136/bmj.e4692

Green, D., Schneider, M., Griswold, H., Belfi, B., Herrera, M., & DeBlasi, A. (2016). A comparison of the HCR-20 V3 among male and female insanity acquittees: A retrospective file study. International Journal of Forensic Mental Health, 15(1), 48–64. https://doi.org/10.1080/14999013.2015.1134726

Hanson, R. K., Helmus, L., & Bourgon, G. (2007). The validity of risk assessments for intimate partner violence: A meta-analysis. Public Safety Canada. https://www.publicsafety.gc.ca/cnt/rsrcs/pblctns/ntmt-prtnr-vlnce/index-en.aspx

Hart, S. D., Douglas, K. S., & Guy, L. S. (2016). The structured professional judgement approach to violence risk assessment. In D. P. Boer (Ed.), The Wiley handbook on the theories, assessment and treatment of sexual offending (pp. 643–666). Wiley. https://doi.org/10.1002/9781118574003.wattso030

Hart, S. D., & Logan, C. (2011). Formulation of violence risk using evidence-based assessments: The structured professional judgment approach. In P. Sturmey & M. McMurran (Eds.), Forensic case formulation (pp. 81–106). Wiley-Blackwell. https://doi.org/10.1002/9781119977018.ch4

Holtzworth-Munroe, A., & Stuart, G. L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological Bulletin, 116(3), 476–497. https://doi.org/10.1037/0033-2909.116.3.476

Kötter, S., von Franqué, F., Bolzmacher, M., Eucker, S., Holzinger, B., & Müller-Isberner, R. (2014). The HCR-20 V3 in Germany. International Journal of Forensic Mental Health, 13(2), 122–129. https://doi.org/10.1080/14999013.2014.911784

Nijdam-Jones, A., García-López, E., Merchan-Rojas, L., Ruiz Guarneros, A., & Rosenfeld, B. (2021). Predictive validity of the HCR-20 V3 with incarcerated males in Mexico City. Criminal Justice and Behavior, 48(10), 1450–1467. https://doi.org/10.1177/0093854821997520

Rossdale, S.-V., Tully, R. J., & Egan, V. (2020). The HCR-20 for predicting violence in adult females: A meta-analysis. Journal of Forensic Psychology Research and Practice, 20(1), 15–52. https://doi.org/10.1080/24732850.2019.1681875

Singh, J. P., Grann, M., & Fazel, S. (2011). A comparative study of violence risk assessment tools: A systematic review and metaregression analysis of 68 studies involving 25,980 participants. Clinical Psychology Review, 31(3), 499–513. https://doi.org/10.1016/j.cpr.2010.11.009

Singh, J. P., Grann, M., & Fazel, S. (2013). Authorship bias in violence risk assessment? A systematic review and meta-analysis. PLoS ONE, 8(9), e72484. https://doi.org/10.1371/journal.pone.0072484

Strub, D. S., Douglas, K. S., & Nicholls, T. L. (2014). The validity of Version 3 of the HCR-20 violence risk assessment scheme amongst offenders and civil psychiatric patients. International Journal of Forensic Mental Health, 13(2), 148–159. https://doi.org/10.1080/14999013.2014.911785

Wolf, V., Mayer, J., Steiner, I., Franke, I., Klein, V., Streb, J., & Dudeck, M. (2024). Predicting violence in female forensic inpatients with substance use disorders – the utility of a gender-responsive assessment. Frontiers in Psychiatry, 15, 1346815. https://doi.org/10.3389/fpsyt.2024.1346815

Legal Authorities

Barefoot v. Estelle, 463 U.S. 880 (1983).

Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993).

Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999).

Simpson v. Miller, 241 Ariz. 341, 387 P.3d 1270 (2017).

State v. Haskie, 242 Ariz. 582, 399 P.3d 657 (2017).

State v. Ketchner, 236 Ariz. 262, 339 P.3d 645 (2014).

United States v. Houser, 36 M.J. 392 (C.M.A. 1993).

Ariz. Rev. Stat. Ann. § 13-3967.

Ariz. Rev. Stat. Ann. § 25-403.03 (as amended by 2026 Ariz. Sess. Laws ch. 263 (H.B. 2995)).

Ariz. Sup. Ct. Order R-23-0004 (2023) (amending Ariz. R. Evid. 702, effective Jan. 1, 2024).

Fed. R. Evid. 404(b).

Fed. R. Evid. 702; Mil. R. Evid. 702; Ariz. R. Evid. 702.

Related topic papers: IPV Risk Assessment Instruments · Borderline Personality and Intimate Partner Violence · Co-Occurrence of IPV and Child Maltreatment · Child Physical Abuse Risk Assessment · Protective Orders and Custody