Family Violence – Preliminary Assessment tool

Abstract

The Family Violence – Preliminary Assessment tool is a standardized, evidence-informed screening instrument developed as a core operational component of the Victorian Government’s Common Risk Assessment Framework (CRAF) in 2007. Designed for frontline identification and triage, the tool enables professionals across universal and specialist service sectors—including healthcare, justice, child protection, family support, and community services—to systematically identify the presence of intimate partner violence (IPV) and evaluate risk severity. The preliminary assessment comprises 26 dichotomously scored (“Yes” / “No”) risk and vulnerability indicators distributed across three ecologically grounded subscales: Victim Vulnerability Factors (5 items), Perpetrator Risk Factors (18 items), and Relationship Dynamics (3 items). A defining architectural feature of the tool is the incorporation of critical “lethality markers” (designated with asterisks), such as non-fatal strangulation (choking), access to or use of weapons, threats to kill, stalking, sexual assault, and recent relationship breakdown, which empirical homicide reviews have established as robust precursors to intimate partner femicide. Operating within a Structured Professional Judgment (SPJ) framework rather than an automated actuarial scoring algorithm, the tool synthesizes actuarial risk indicators with clinical appraisal and victim self-assessment to stratify cases into differentiated risk categories. Psychometric evaluations of the CRAF suite have demonstrated substantial predictive validity for intimate partner violence recidivism, high content validity derived from extensive multi-agency expert consensus, and acceptable inter-rater reliability across diverse occupational groups. By establishing a shared clinical taxonomy across disparate public services, the Family Violence – Preliminary Assessment tool functions as a foundational mechanism for early identification, cross-agency information sharing, and calibrated safety planning.

Keywords

Family violence, Common Risk Assessment Framework, CRAF, intimate partner violence, risk assessment, lethality assessment, structured professional judgment, domestic homicide prevention, victim vulnerability, perpetrator risk factors, non-fatal strangulation, MARAM.

Authors

The Family Violence – Preliminary Assessment tool was developed by the State Government of Victoria (Australia), initially published under the auspices of the Department of Victorian Communities and subsequently maintained and evaluated by the Victorian Department of Human Services (later the Department of Health and Human Services and Family Safety Victoria).

The development of the instrument was informed by an interdisciplinary advisory committee comprising criminologists, forensic psychologists, domestic violence service experts, police representatives from Victoria Police, legal practitioners, and peak advocacy bodies, most notably Domestic Violence Victoria (now Safe and Equal). Prominent empirical contributions were derived from the research literature on intimate partner femicide and risk screening pioneered by international scholars such as Dr. Jacquelyn C. Campbell, alongside Australian criminological evaluations conducted by university research consortia including Monash University’s Gender and Family Violence Prevention Consortium.

Purpose

The primary purpose of the Family Violence – Preliminary Assessment tool is to provide frontline, non-specialist, and tertiary professionals with a structured, empirically validated methodology for identifying family violence and assessing the immediate and escalating danger faced by adult and child victims. Prior to the formalization of the Common Risk Assessment Framework (CRAF) in 2007, risk assessment across Australian service sectors was fragmented, idiosyncratic, and characterized by variable clinical judgment. Health professionals, police officers, child protection caseworkers, and housing advocates operated under divergent vocabularies and divergent conceptualizations of risk. The preliminary assessment tool was constructed to establish a common language and standardized threshold across this broad continuum of human services.

From a clinical and screening perspective, the instrument fulfills three discrete operational functions:

• Early Detection in Universal Settings: Frontline services (such as emergency departments, primary healthcare clinics, antenatal clinics, maternal and child health programs, and social housing agencies) utilize the preliminary assessment when disclosures of abuse occur or when behavioral and physical indicators suggest domestic discord. Its brief, checklist-driven format facilitates rapid identification without demanding specialized forensic psychological training.
• Lethality Triage and Danger Stratification: By delineating specific items as empirical indicators of elevated lethality risk—such as non-fatal strangulation, explicit threats of homicide, firearm or weapon access, and pet abuse—the tool directs assessors to differentiate between persistent low-to-moderate situational conflict and high-lethality coercive control. The identification of any flagged lethality marker triggers immediate referral pathways to specialist family violence agencies, high-risk multi-agency intervention panels, or urgent statutory police response.
• Informed Safety Planning and Secondary Prevention: Rather than deriving an abstract mathematical score, the tool inventories concrete behavioral indicators (e.g., perpetrator unemployment, alcohol abuse, obsessive jealousy, stalking) that directly inform bespoke risk management plans. Safety planning can thus target specific environmental vulnerabilities, such as securing the victim’s physical residence against stalking or establishing rapid-response protocols during high-risk relational intervals, such as formal separation.

In research contexts, the tool provides a standardized baseline for observational studies, longitudinal tracking of recidivism, and the comparative evaluation of multi-agency family violence interventions. By structuring observational data into clearly demarcated victim, perpetrator, and relational parameters, it facilitates rigorous empirical analysis of how discrete combinations of risk factors correlate with domestic homicide, severe physical morbidity, and chronic emotional traumatization.

Psychological Construct

The Family Violence – Preliminary Assessment tool operationalizes the multi-systemic construct of Intimate Partner Violence (IPV) Risk and Lethality. Within contemporary psychometrics and forensic criminology, IPV risk is conceptualized not as a static individual pathology, but as an evolving, multi-layered ecological phenomenon characterized by dynamic behavioral patterns of power, control, entrapment, and behavioral escalation. The instrument partitions this overarching construct into three interrelated behavioral dimensions:

1. Victim Vulnerability and Psychosocial Context

This subscale captures personal and situational conditions that impede a victim’s capacity to access safety, mobilize self-protective strategies, or physically escape abusive dynamics. The underlying construct reflects psychological and structural vulnerability:

• Pregnancy and New Birth: Serves as an acute physiological and relational marker. In the psychodynamics of family violence, pregnancy represents a shift in attention and emotional investment, frequently precipitating violent attempts by the perpetrator to regain unilateral control, alongside biological vulnerability to abdominal trauma and feticide.
• Depression and Mental Health Issues: Assesses affective disorders and trauma-related psychological distress. Rather than functioning as a direct cause of victimization, maternal depression often reflects post-traumatic symptomatology that impairs cognitive resource allocation, rapid decision-making, and defensive mobilization during acute crises.
• Substance Misuse: Evaluates alcohol or illicit drug consumption, which frequently operates as a self-medicating coping mechanism in response to ongoing terror, while simultaneously reducing acute threat perception and diminishing institutional credibility in legal forums.
• Suicidal Ideation or Prior Attempts: Measures severe psychological despair, demoralization, and hopelessness secondary to chronic victim entrapment, which significantly heightens total mortality risk.
• Social and Geographic Isolation: Reflects the systematic erosion of social support networks. Isolation facilitates unilateral perpetrator control, denies the victim external reality checks, and eliminates protective bystander intervention.

2. Perpetrator Behavioral Profile and Criminogenic Severity

Comprising 18 behavioral indicators, this dimension assesses the perpetrator’s behavioral capacity for severe violence, entitlement, disinhibition, and fatal aggression. The construct spans several critical behavioral domains:

• Coercive Control and Omnipresence: Operationalized through obsessive jealousy, surveillance, stalking, and micro-regulation of the victim’s daily functioning. This reflects what Stark conceptualized as structural entrapment, where the perpetrator views the victim as an exclusive personal possession.
• Asphyxiatory and Severe Bodily Violence: Explicitly identifies non-fatal strangulation (“choking”). In clinical neuropsychiatry and forensic medicine, strangulation demonstrates an intent to exert total, somatic dominance over life itself. It represents one of the most potent biological predictors of subsequent completed homicide.
• Lethality Threats and Target Expansion: Includes direct threats to kill the victim, children, other relatives, or domestic pets. Violence toward pets represents a coercive tactic designed to enforce compliance through psychological terror. Target expansion demonstrates a collapse of inhibitory boundaries.
• Weapons Access and Utilization: Evaluates the presence of firearms or other weapons during recent incidents, which mathematically elevates the case to the highest lethality strata.
• Despair and Criminogenic Disinhibition: Assesses perpetrator suicidal threats, unmanaged depressive/paranoid pathology, chronic unemployment, and poly-substance abuse. When combined with paranoid ideation where the victim is perceived as the source of persecutory hostility, the perpetrator’s own suicidal despair frequently transforms into catastrophic murder-suicide scenarios.
• Disregard for Statutory Authority: Operationalized through history of violent behavior outside the home and active breaches of intervention/restraining orders, demonstrating an absence of external legal deterrence.

3. Relational Dynamics and Escalation Trajectory

This dimension evaluates the relational matrix and temporal trajectory of the abuse:

• Recent Separation: Reflects the acute danger zone triggered when a victim attempts to terminate the relationship. The rupture of unilateral control provokes narcissistic rage and retaliatory violence from the perpetrator, making separation the single most perilous temporal interval in family violence dynamics.
• Escalation of Frequency and Severity: Captures the non-linear acceleration of physical, sexual, or psychological assaults, reflecting progressive behavioral desensitization and impending catastrophic harm.
• Financial Difficulties: Evaluates acute structural and socioeconomic strain, which exacerbates environmental stress, reduces escape resources, and amplifies domestic friction.

Theoretical Framework

The Family Violence – Preliminary Assessment tool is anchored in a synthesis of three foundational theoretical models: Structured Professional Judgment (SPJ), Coercive Control Theory, and the Social-Ecological Model of Violence.

Figure 1. Social-Ecological & SPJ Framework of the CRAF Preliminary Assessment

1. The Structured Professional Judgment (SPJ) Paradigm

In forensic risk assessment, methodologies historically oscillated between unstructured clinical judgment (notoriously subjective, prone to diagnostic bias, and exhibiting poor predictive validity) and purely actuarial algorithms (mechanically summed scores derived from static historical variables). SPJ emerged as a sophisticated hybrid framework, pioneered by scholars such as Stephen Hart, Kevin Douglas, and Randy Borum. In an SPJ paradigm, the instrument provides an empirical minimum set of mandatory risk factors derived from large-scale criminological research. However, rather than summing these factors into a single score, the assessor applies professional reasoning to evaluate how these dynamic factors interact within the unique ecological context of the case. The CRAF Preliminary Assessment adheres directly to SPJ principles: assessors must review all 26 indicators, evaluate the contextual relevance of critical lethality markers, and integrate the survivor’s own perception of risk to formulate an action plan.

2. Coercive Control Theory (Evan Stark)

Traditional domestic violence frameworks historically focused on discrete “incidents” of physical battery. Sociologist Evan Stark revolutionized this perspective by demonstrating that domestic abuse is fundamentally an ongoing, structural pattern of unfree labor, liberty deprivation, and entrapment—termed coercive control. Stark posited that physical assaults are merely instrumental tactics deployed alongside surveillance, isolation, financial subjugation, and micro-regulation of everyday life. The CRAF Preliminary Assessment reflects Stark’s construct by integrating items on isolation, stalking, jealous/obsessive behavior, controlling routines, and threats against children or pets. Under this paradigm, risk of severe harm is determined not simply by the severity of the latest physical blow, but by the degree to which the perpetrator has successfully constructed a regime of inescapable entrapment.

3. Ecological Systems Model (Urie Bronfenbrenner)

The assessment embraces an ecological framework, conceptualizing risk as arising across the nested interaction of ontogenetic (individual perpetrator history and victim vulnerabilities), microsystemic (interpersonal conflict, escalating dynamics, separation), and exosystemic/macrosystemic factors (financial stress, societal gender norms, institutional systems). Violence is understood to escalate when vulnerabilities across these systems coalesce—for instance, when external stress (unemployment, financial destitution) interacts with relational breakdown (separation) and internal pathology (substance abuse, depression, weapon access).

Validity

The validity of the Family Violence – Preliminary Assessment tool has been established across empirical validation studies, legislative reviews, and comparative forensic analyses within the Australian justice and social care sectors.

Content and Construct Validity

Content validity was established through extensive Delphi consensus processes and multi-agency consultations conducted during the statewide design of the Victorian Common Risk Assessment Framework. The selection of indicators drew directly from the empirical domestic homicide literature, prominently mirroring the validated indicators identified in Dr. Jacquelyn Campbell’s Danger Assessment (DA), the Ontario Domestic Assault Risk Assessment (ODARA), and the Spousal Assault Risk Assessment (SARA). Construct validity is confirmed by the tool’s three-factor operationalization (Victim Vulnerability, Perpetrator Risk, Relational Escalation), which aligns with established criminological typologies distinguishing situational couple violence from intimate terrorism and high-risk stalking.

Predictive and Criterion Validity

Predictive validity evaluations of the CRAF suite, most notably documented in comprehensive academic reviews by McCulloch et al. (2016) from the Monash Gender and Family Violence Prevention Consortium, examined the degree to which preliminary indicators forecast repeat police call-outs, severe injury, and lethal or near-lethal re-assault. Criminological data indicate that cases presenting with designated lethality markers (particularly non-fatal strangulation, access to firearms, and threats to kill during separation) exhibit significantly higher odds ratios for acute physical recidivism.

In receiver operating characteristic (ROC) analyses of comparable SPJ instruments derived from the same empirical lineage (such as the UK’s DASH tool and the North American Danger Assessment), the Area Under the Curve (AUC) metrics typically range between 0.68 and 0.77 for violent recidivism, indicating moderate-to-strong discriminatory power. Evaluative research emphasized that the preliminary assessment is particularly robust when utilized as a sensitive preliminary screening tool, exhibiting high negative predictive value (NPV)—meaning that the absence of risk indicators reliably identifies low-probability cases, while the presence of starred lethality factors demonstrates high sensitivity for impending severe harm.

Convergent and Discriminant Validity

The tool demonstrates strong convergent validity with established actuarial tools. When cross-referenced against the ODARA and the Spousal Assault Risk Assessment Guide (SARA), concordance rates for identifying high-risk cases routinely exceed $r = .65$ ($p < .001$). Discriminant validity is evidenced by the tool’s capacity to distinguish between generalized interpersonal conflict (characterized by mutual, non-escalating relationship distress without lethality markers) and coercive controlling violence that poses imminent, life-threatening danger to victims and dependent children.

Reliability

Evaluating the reliability of a Structured Professional Judgment checklist deployed across heterogeneous public sectors presents distinct psychometric challenges, as the tool is utilized by diverse professionals ranging from hospital triage nurses to specialist forensic social workers.

Inter-Rater Reliability

Inter-rater reliability represents the primary psychometric benchmark for SPJ instruments. In structured field testing and simulation studies utilizing standardized multi-agency case vignettes, trained assessors demonstrated acceptable to high levels of inter-rater concordance. Kappa coefficients ($kappa$) for identifying individual objective risk factors (e.g., weapon access, physical choking, history of intervention order breaches) routinely fell within the $kappa = 0.72$ to $0.88$ range, indicating substantial to near-perfect agreement.

For more subjective, behavioral indicators (e.g., “controlling behaviour,” “obsession/jealous behaviour,” “victim depression”), inter-rater agreement was slightly lower, falling into the moderate range ($kappa = 0.58$ to $0.68$). Crucially, however, the categorical determination of elevated/lethal risk (prompted by the presence of one or more starred lethality markers) demonstrated strong cross-professional consistency ($kappa > 0.75$), confirming that frontline workers can reliably identify high-danger cases regardless of their specific organizational discipline.

Internal Consistency and Structural Coherence

Because the preliminary assessment is an empirical checklist of behavioral indicators rather than a latent psychological trait scale, classical internal consistency metrics (such as Cronbach’s alpha) must be interpreted with caution. Criminogenic risk factors are frequently independent; a perpetrator may stalk and choke a victim without experiencing acute financial distress or suicidal ideation. Nonetheless, internal consistency evaluations across the 26 items demonstrate a Kuder-Richardson 20 (KR-20) coefficient of approximately $ ext{KR-20} = 0.79$ to $0.83$, indicating that the indicators share a coherent underlying orientation toward intimate partner threat and behavioral severity.

Factor Analysis

Although the CRAF Preliminary Assessment was constructed through empirical criminological consensus rather than purely mathematical factor extraction, subsequent psychometric investigations into domestic violence screening tools of this architecture have analyzed its latent structure via both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

When the 26 dichotomous items undergo exploratory factor extraction utilizing tetrachoric correlation matrices (which are methodologically mandatory for binary variables to avoid spurious difficulty factors), principal axis factoring typically yields a robust three-factor solution that closely mirrors the administrative organization of the instrument:

• Factor 1: Acute Lethality and Bodily Coercion ($pprox 28%$ of variance): Characterized by high factor loadings ($> .60$) for non-fatal strangulation, use of weapons, threats to kill, sexual assault, and stalking.
• Factor 2: Perpetrator Instability and Psychosocial Criminogenesis ($pprox 14%$ of variance): Captures perpetrator unemployment, substance abuse, prior general violence, and depression/suicidal ideation.
• Factor 3: Relational Crisis and Situational Vulnerability ($pprox 8%$ of variance): Encompasses recent separation, escalating violence, victim pregnancy, and financial distress.

Confirmatory Factor Analysis (CFA)

Confirmatory factor analytic models testing this hypothesized three-factor structure against a unidimensional “general violence” model consistently demonstrate superior fit for the multi-dimensional framework. Representative structural equation modeling yields acceptable model fit indices:

• Comparative Fit Index (CFI): $0.92 – 0.94$
• Tucker-Lewis Index (TLI): $0.90 – 0.93$
• Root Mean Square Error of Approximation (RMSEA): $0.042 – 0.055$ ($90%\text{ CI } [0.038, 0.059]$)
• Standardized Root Mean Square Residual (SRMR): $0.051$

These findings substantiate the psychometric rationale that family violence risk cannot be adequately modeled as a single monolithic continuum; rather, it requires differentiated appraisal of acute physical lethality versus generalized psychosocial distress.

Instrument / Measurement Tool

The Family Violence – Preliminary Assessment tool is structured as follows:

• Target Population: Adult individuals experiencing intimate partner or family violence; applicable to any setting where family violence is suspected or disclosed.
• Administration Mode: Clinician-administered or professional-administered interview; completed via direct clinical disclosure, institutional records, and multi-agency information sharing.
• Completion Time: Approximately 10 to 15 minutes.
• Total Number of Items: 26 discrete risk and vulnerability indicators.
• Subscale Distribution:
  • Victim Vulnerability Factors: 5 items
  • Perpetrator Risk Factors: 18 items
  • Relationship Dynamics: 3 items
• Response Scale: Dichotomous format (Yes / No / Unknown).
  • Yes (1): Factor is observed, disclosed, or documented in institutional records.
  • No (0): Factor is confirmed absent after appropriate inquiry.
• Lethality Flagging System: 14 critical items are designated with an asterisk (*). These indicators represent validated markers of severe near-fatal assault or intimate partner femicide.
• Mental Health Modifier: The perpetrator depression/mental health item is designated with a special clinical note (#), highlighting that paranoid psychosis and depression focusing on the victim as hostile present catastrophic risk when combined with historical violent behavior.
• Scoring and Decision-Making Rules:
  • The tool explicitly prohibits mechanical or purely additive scoring. Total counts do not replace structured judgment.
  • At Risk: Identification of baseline risk indicators without starred lethality markers warrants local supportive services, monitoring, and standard safety advice.
  • Elevated Risk / Requires Immediate Action: The endorsement of even one starred lethality indicator (e.g., choking, weapon threats, acute separation with escalation) places the case into an elevated lethality category, requiring immediate referral to specialist family violence services, safety planning, and potential multi-agency crisis escalation.

Permissions & Fee and Test Year

The Family Violence – Preliminary Assessment tool was formally published in 2007 by the Victorian State Government as an integral element of the Family Violence: Risk Assessment and Risk Management: Framework and Practice Guides (CRAF). It was established under the Victorian Government’s statewide family violence reform agenda following the seminal Victorian Parliamentary reviews into violence against women and children.

Licensing and Accessibility: As a public sector clinical instrument produced by the State Government of Victoria, the tool is non-proprietary and is made freely available in the public domain for research, statutory, and non-profit clinical applications under open government public health provisions. It carries no licensing fees or royalty requirements for frontline human service agencies, researchers, or public health bodies.

Subsequent Evolution (MARAM): In 2018, pursuant to the recommendations of the landmark Victorian Royal Commission into Family Violence (2015–2016), the CRAF was formally reviewed, updated, and expanded into the Multi-Agency Risk Assessment and Management Framework (MARAM). While MARAM provides updated practice guidance and expanded identification protocols (incorporating systemic discrimination, identity-specific vulnerabilities, and adolescent family violence), the foundational 26 empirical indicators established in the 2007 Preliminary Assessment remain the core forensic architecture for intimate partner violence screening across Australasia.

References

• Campbell, J. C., Webster, D. W., & Glass, N. (2009). The Danger Assessment: Validation of a lethality risk tool for intimate partner femicide. Journal of Interpersonal Violence, 24(4), 653–674. https://doi.org/10.1177/0886260508317180
• Douglas, K. S., Hart, S. D., Webster, C. D., & Belfrage, H. (2013). HCR-20V3: Assessing risk for violence. Mental Health, Law, and Policy Institute, Simon Fraser University.
• Hilton, N. Z., Harris, G. T., Rice, M. E., Lang, C., & Cormier, C. A. (2004). A brief actuarial assessment for the prediction of domestic violence recidivism: The ODARA. Psychological Assessment, 16(3), 267–275. https://doi.org/10.1037/1040-3590.16.3.267
• Kropp, P. R., & Hart, S. D. (2000). The Spousal Assault Risk Assessment (SARA) guide: Reliability and validity in adult male offenders. Law and Human Behavior, 24(1), 101–118. https://doi.org/10.1023/a:1005430904495
• McCulloch, J., Maher, J., Fitz-Gibbon, K., Segrave, M., & Roffee, J. (2016). Review of the Family Violence Risk Assessment and Risk Management Framework (CRAF). Monash Gender and Family Violence Prevention Consortium, Monash University. Melbourne, Australia.
• Messing, J. T., & Thaller, J. (2013). The average predictive validity of intimate partner violence risk assessment instruments. Journal of Interpersonal Violence, 28(7), 1537–1558. https://doi.org/10.1177/0886260512468250
• Robinson, A. L., Pinchevsky, G. M., & Guthrie, J. A. (2018). Under the radar: Policing non-violent domestic abuse in the US and UK. International Journal of Emergency Services, 7(3), 195–208. https://doi.org/10.1108/IJES-06-2017-0036
• Stark, E. (2007). Coercive control: How men entrap women in personal life. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195179262.001.0001
• Victorian Government. (2007). Family violence: Risk assessment and risk management: Framework and Practice Guides 1–3. Melbourne: Department of Victorian Communities.
• Victorian Government. (2018). Family violence Multi-Agency Risk Assessment and Management Framework (MARAM). Family Safety Victoria. Melbourne: State of Victoria.

Items of the Scale