Prevention Insights

The Research Behind Behavioral Risk Assessment: What Changed the Field

August 3, 20264 min readBy Homicide Zero Editorial Team

Research on targeted violence has produced one of its most consequential findings over the past 30 years: attacks are almost never random, and almost never truly impulsive. The U.S. Secret Service National Threat Assessment Center (NTAC), which has studied the behavioral patterns preceding mass casualties and targeted attacks since the 1990s, has found consistently that perpetrators communicate intent, acquire means, and move through recognizable stages before acting. This research changed how professionals think about risk.

The Problem With Profiling

Early approaches to violence prevention focused heavily on identifying a 'type' of person who might commit an attack. Researchers and practitioners tried to build demographic or psychological profiles: certain ages, certain diagnoses, certain backgrounds. The research found this approach does not work. No reliable profile exists. Individuals who commit targeted violence look, on the surface, like many people who never will.

What the research did find was that behavior, not identity, is the reliable signal. Specific, observable patterns of behavior, communication, and circumstance correlate with increased risk in ways that demographic characteristics do not. This finding is the foundation that behavioral risk assessment is built on.

What Behavioral Risk Assessment Measures

Behavioral risk assessment shifts the focus from who someone is to what someone is doing, saying, and experiencing. A structured behavioral risk assessment evaluates factors including recent stressors and grievances, communications about intent or fascination with violence, access to weapons or capacity to cause harm, history of violence or threats, mental health status as it relates to risk (not as a standalone predictor), and the presence or absence of protective factors such as social connection and institutional support.

The goal is not to predict the future with certainty. Behavioral risk assessment produces an estimate of relative risk and, more importantly, identifies which risk factors are present and what interventions might reduce them.

Structured Assessment and Why It Matters

Two broad approaches exist in the field: actuarial models and structured professional judgment. Actuarial models apply fixed scoring to predetermined variables. Structured professional judgment frameworks, which are the basis for most validated threat assessment instruments today, guide clinicians through a structured set of questions while preserving space for clinical interpretation of the whole picture.

Research supports structured approaches over unstructured clinical impression. A clinician making an informal judgment about someone's risk performs less consistently than one using a validated structured instrument. The structure does not replace judgment. It focuses and improves it.

How This Translates to Practice

Organizations applying these research findings typically use a two-step process. A rapid screener, like Homicide Zero's Homicide Threat Screener (HTS), allows non-clinicians to flag and document a concern in a structured way. If the screener indicates elevated risk, a trained clinician follows with a comprehensive behavioral risk assessment, such as the Homicide Safety Risk Assessment (HSRA), which evaluates the full range of risk and protective factors and guides next steps.

Neither tool attempts to predict certainty. Both translate the research base into a structured process that turns a concern into a reviewable, actionable evaluation.

Where to Go From Here

The field of behavioral risk assessment continues to evolve as researchers refine what works. Organizations building or improving their threat assessment capacity can start with the research base: NTAC's published reports, FBI threat assessment resources, and peer-reviewed studies on structured professional judgment are publicly available and provide useful grounding before selecting any tool or protocol.