Static and Dynamic Risk Factors: What They Mean in Forensic Psychology
Author: Ukrainian Psychological Hub · Published: October 11, 2026 · Editorial Policy
Static and dynamic risk factors are categories used in forensic psychology to organize information associated with future violence or reoffending. Static factors describe an individual's relevant history or other features that cannot be changed by current intervention. Dynamic factors describe present characteristics or circumstances that may change over time. Together they can inform structured assessments, supervision, treatment planning, and review of changing circumstances. They do not establish that someone will commit an offense, that a diagnosis causes violence, or that a risk score determines a legal outcome.
The distinction answers two different questions. Historical information can help estimate patterns of risk observed in comparable groups. Current and potentially modifiable information can help identify what requires attention now and what might warrant reassessment later. The evidentiary challenge is to establish which factors predict which outcomes, in which populations and settings, over what period, and whether observed change actually improves decisions or outcomes. Reviews by Ogonah and colleagues (2023) and Fazel and colleagues (2024) show why every part of that chain deserves scrutiny.
What Are Static and Dynamic Risk Factors?
A forensic risk factor is a measured characteristic, circumstance, or historical event statistically associated with a specified adverse outcome in an identified population. The outcome might be violent reoffending, any new conviction, a sexual offense, institutional aggression, or another clearly defined event. The same item can have different relevance across outcomes. A correlation with arrest or reincarceration is not automatically a correlation with violent behavior, since policing, supervision, charging practices, and legal processes also influence those recorded outcomes.
A static risk factor is usually fixed historical information. Prior convictions, documented previous violence, age at first recorded offense, and an established history of supervision violations are familiar examples. An assessor can investigate whether the history has been recorded accurately, but cannot rewrite the past. Static instruments sometimes use time-related variables or derived estimates that can change as a person ages, so 'static' describes the broad evidential role of a factor rather than an absolute promise that every number remains identical forever. Public Safety Canada's research summary on static and dynamic approaches makes this practical distinction explicit.
A dynamic risk factor is a feature of current functioning or environment that may vary. Depending on the instrument and outcome studied, relevant examples can include ongoing substance-related difficulties, attitudes that support offending, current associations, engagement with supervision, acute conflict, or instability in accommodation. Changeability alone does not establish predictive validity. 'Unemployment,' for example, has different meanings across labor markets and disabilities; 'noncompliance' can reflect access barriers as well as individual behavior. A factor deserves forensic weight only when its definition, measurement, context, and evidence support that use.
Static vs. Dynamic Risk Factors: The Core Differences
Static factors give an assessment a historical reference point. Dynamic factors invite a time-stamped appraisal of current circumstances and possible management responses. Static predictors can retain statistical value long after the original event; dynamic indicators require repeated observation and careful interpretation. Neither category is intrinsically more scientific, and neither can replace a clearly stated outcome definition, a suitable comparison population, and transparent assumptions.
The most consequential difference is practical. A treatment program cannot change someone's previous convictions, but it may be able to support stable housing, address an identified substance-use problem, improve access to services, or help a person develop alternatives to offending-related behavior. Yet an improvement in housing or treatment attendance is not, by itself, proof that the probability of serious violence has fallen. It is evidence that one circumstance has changed, and the interpretation depends on what has been validated.
Forensic professionals therefore distinguish the act of observing a change from the inference that a person's future outcome has changed. A rating may decline because the person is in a safer environment, because an assessor has obtained better information, because opportunities for offending have been restricted, or because the underlying characteristic truly has changed. An assessment must specify which explanation is supported. The NIJ reentry study found that adding rated dynamic factors did not improve prediction beyond static variables in its study setting, illustrating why the additional value cannot be presumed.
Examples of Static Risk Factors in Forensic Assessments
Prior violent behavior is often considered because it is directly relevant to some subsequent violence outcomes. That does not make every prior offense equivalent or guarantee repetition. Date, severity, setting, opportunity, circumstances, elapsed time, and quality of records all matter. A single event documented in an administrative file is a historical fact requiring interpretation, not a psychological diagnosis. Convictions and arrests also differ as indicators: arrest records reflect criminal-justice contact and are shaped partly by enforcement and reporting.
Age at first offense or first documented justice contact appears in several historical risk frameworks. Its association with later outcomes is a group-level finding, complicated by age, exposure, recording practices, developmental context, and the outcome being predicted. A record of childhood adversity can likewise be historically fixed, but it is neither an individual forecast nor a moral characteristic. Experiences of maltreatment should not become automatic labels of dangerousness; the relevance of any such measure requires specific empirical support.
Other historical variables might include previous failed supervision, documented repeated rule violations, and patterns of past behavior directly connected to the referral question. Some tools include demographic variables, raising particular questions of fairness, necessity, and legal acceptability. A person may also age into a different statistical reference category while retaining an unchanged offense history. The examiner should distinguish the underlying historical record from an age-adjusted estimate derived by a model.
Examples of Dynamic Risk Factors
Stable Dynamic Factors
Stable dynamic factors are characteristics or circumstances that tend to persist for months or years but can change with sustained intervention or changes in life conditions. In particular correctional and sexual-recidivism research, examples may include enduring attitudes favorable to offending, persistent difficulties in regulating certain behaviors, long-term associations, or recurring relationship patterns. These are measured constructs defined by particular instruments. They are not free-floating descriptions that allow an observer to score someone's character by intuition.
A stable factor can remain present during a relatively peaceful interval. Conversely, an intervention may produce changes in self-reported attitudes without producing an independently observed reduction in reoffending. Studies of dynamic instruments in sexual-offense populations have found predictive associations and some additional predictive value beyond static instruments, but effects and performance vary. The meta-analysis indexed in PubMed is an important source for incremental validity and change-score findings; those results should be read within the specialized populations that were studied.
Acute Dynamic Factors
Acute dynamic factors fluctuate over shorter periods, sometimes days or weeks. Depending on the outcome and case formulation, a sudden breakdown in housing arrangements, an emerging conflict, an escalation in substance use, changes in access to a potential target, or a breakdown in supervision may prompt immediate review. A temporary change is not automatically evidence of imminent violence. The relevance must rest on observations, timing, context, and available research on the specific factor.
The distinction between stable and acute factors comes from work that tries to connect longer-term vulnerabilities with shorter-term shifts. It is especially important to avoid treating every episode of anger, intoxication, stress, or interpersonal difficulty as an impending offense. Most people who experience these conditions do not engage in criminal violence. The question for a forensic professional is how an observed condition relates to a defined concern for this person, in this environment, with this history, using a sufficiently tested assessment approach.
Dynamic Factors Are Not the Same as Criminogenic Needs
A criminogenic need is generally a dynamic domain selected for intervention because of its empirical association with reoffending within a correctional model. The risk–need–responsivity approach uses the language of risk to estimate relative need for services, need to identify intervention targets, and responsivity to adapt delivery to an individual's abilities and circumstances. The concept has operational value, but a label used in a program is not proof that the factor directly causes an offense or that reducing a score is sufficient to change criminal behavior.
The 2024 umbrella review of risk–need–responsivity evidence synthesized 26 meta-analyses and found mixed support, substantial methodological limitations, and often low-quality evidence. Some studies show associations between program characteristics and lower recidivism, while others do not. The review's central lesson is methodological: program adherence, risk classification, proposed mechanisms, and actual reoffending outcomes must each be evaluated separately. No single domain supplies a universal account of why crime occurs.
A dynamic risk item may bundle several processes under one label. 'Relationship instability' could involve conflict, partner behavior, economic dependence, housing access, or stress. 'Antisocial attitudes' can depend strongly on how interview questions are asked and interpreted. Treating such broad labels as direct mechanisms obscures the path from evidence to inference. A review of conceptualization and measurement asks whether dynamic factors are genuinely changeable, whether observed changes forecast outcomes, and whether targeted interventions have demonstrated the predicted consequences.
Protective Factors: A Related but Distinct Idea
Protective factors are personal, interpersonal, or environmental conditions associated with lower levels of an outcome, or capable of supporting effective risk management. Examples can include dependable community support, engagement with meaningful activities, accessible treatment, and a practical safety plan. Their value must be demonstrated rather than inferred from good intentions or conventional expectations about a 'successful life.' A protective condition can coexist with substantial historical risk, and an individual may have vulnerabilities alongside significant strengths.
The presence of protection does not simply cancel out a historical factor in an arithmetic sum. Instruments differ in how they measure strengths, and interactions among factors may depend on a person's environment. An assessment that reports only problems can miss opportunities to make interventions effective, while an assessment that lists strengths without testing their relevance can offer false reassurance. Protective factors should be connected to observable behaviors, available resources, and a defined management plan.
How Forensic Psychologists Assess These Factors
Assessment begins with a referral question: risk of what, over what time frame, for what decision? Violence within a locked ward over the next week, reoffending after release over two years, and breach of community-supervision conditions are distinct outcomes. A valid tool for one cannot automatically be transported to the others. The legal decision-maker also needs to know whether a report supplies psychological evidence, a conditional risk formulation, or a proposition that the court must decide under local law.
A competent evaluation uses multiple sources as appropriate: relevant records, direct interview, collateral information where ethically and legally available, behavioral observations, and assessment methods supported for the population and purpose. The APA Specialty Guidelines for Forensic Psychology emphasize scientific foundations, appropriate methods, relevant individual differences, limitations, and clear documentation. Records can be incomplete or contradictory; interview responses can be influenced by incentives and stress; and observations within an institution may not transfer readily to community settings.
Static information often requires verifying the date and nature of each historical item. Dynamic information requires operational definitions and assessment dates. A sentence such as 'risk increased because treatment motivation declined' is incomplete unless the reviewer knows how motivation was assessed, how much it changed, what alternatives could explain that change, and whether the change meaningfully predicts the specified outcome. Repeated ratings are most useful when the instrument has acceptable reliability and evidence for longitudinal interpretation.
For a broader account of methods, confidentiality, record review, and the role of legal instructions, see our forensic psychological assessment overview. Static and dynamic factors are components of that process. They never replace professional competence, information about the case, or the legal framework governing how an opinion may be used.
Structured Tools: What They Actually Do
Forensic practitioners may organize risk information through actuarial methods, structured professional judgment, or structured risk-and-needs frameworks. Actuarial tools use predefined combinations of variables to estimate statistical risk groups or probabilities for specified outcomes. Structured professional judgment approaches identify relevant factors and guide a documented formulation about plausible scenarios and management. These approaches can overlap in content while differing in how a final judgment is generated.
The HCR-20 family, for instance, explicitly organizes information into historical, clinical, and risk-management domains. Historical items contribute a background account, while clinical and prospective management items invite attention to change and context. That architecture is not itself validation of the precision of a forecast; different versions and study settings must be considered separately. The 2026 systematic review of longitudinal HCR-20 properties found that dynamic scores can decline during treatment but offered little evidence that changes in scores predict recidivism.
Static-99 is a name for a specialized actuarial instrument family for sexual-recidivism assessment, not a generic 'static factor checklist' for any violence. A study of Static-99 in a forensic hospital sample does not establish its performance in every correctional or clinical population. Similarly, a psychopathy measure is a construct assessment and should not be presented as a verdict, a comprehensive violence forecast, or a clinical diagnosis equivalent to criminality. Instrument choice belongs to the precise referral question.
What Does Research Say About Predictive Accuracy?
Risk research commonly reports discrimination: how well an instrument tends to rank people who later experience an outcome against those who do not. The area under the receiver operating characteristic curve, or AUC, summarizes that ranking performance. An AUC of 0.50 corresponds to chance ranking under the usual definition; a higher AUC indicates better average discrimination in the particular sample. It does not directly state an individual's probability of violence, and it does not tell us whether predicted probabilities are calibrated.
In Ogonah and colleagues' 2023 systematic review and meta-analysis, 50 publications covering 10,460 people discharged from forensic psychiatric hospitals assessed 36 tools. Estimates for pooled AUCs of selected measures of violent recidivism ranged approximately from 0.64 to 0.72. Nearly all included studies were rated at high risk of bias, and calibration or full error metrics were seldom reported. These figures describe an evidence base in specialist forensic hospital discharge contexts; they are not universal performance claims for ordinary defendants, juveniles, or all people with mental-health conditions.
The review found that 88% of participants were men and all included research came from high-income countries. A predictor that works modestly in one system may operate differently where access to services, legal definitions, follow-up durations, or recorded outcomes differ. Instruments can also be used outside the purpose for which they were designed, reducing the relevance of the validation evidence. The combination of statistical association and institutional routine is not equivalent to proof of clinical or legal correctness.
Does Adding Dynamic Factors Improve Prediction?
Sometimes dynamic measures offer additional information beyond historical scores, especially where the assessment was designed to capture relevant change. But the incremental benefit is an empirical question. It must be tested against a defensible comparison model, ideally in an independent sample, with clear outcomes and sufficient follow-up. A statistically significant increment may still be too small to improve decisions or justify the burden of repeated assessment.
A meta-analysis of dynamic instruments for sexual-recidivism assessment drew on 52 studies and reported predictive associations, modest incremental effects beyond static instruments, and findings for change scores. Its population and outcomes were specialized; it does not license generalizing to every kind of violent or nonviolent offending. A separate NIJ-funded reentry study did not find incremental predictive benefit from rated dynamic measures beyond static variables, while some self-reported information was predictive. The results are compatible with dynamic assessment being useful in some contexts and disappointing in others.
These findings should also be distinguished from causal intervention trials. A factor can improve a prediction because it is a marker of broader conditions, without necessarily being a productive treatment target. A dynamic score could change after intervention while other factors, access to opportunity, or reporting practices are responsible for subsequent outcomes. The evidence needed to claim 'changing X reduces violence' is stronger than the evidence needed to say 'X is associated with violence in a validation sample.'
Can a Falling Dynamic Score Prove That Risk Has Fallen?
A lower dynamic score establishes that a measured indicator or composite has changed under the instrument's procedures. To conclude that the probability of a specified outcome has also fallen, the reviewer needs evidence that within-person changes on the instrument predict later outcomes, beyond baseline risk and regression to the mean. They must also ask whether change exceeds measurement error, whether raters agree, and whether the original measure was sensitive to the kind of improvement being claimed.
The 2026 review of longitudinal HCR-20 research is instructive because it differentiates declining ratings during treatment from the limited evidence linking those changes to later recidivism. A service can record therapeutic progress accurately while remaining uncertain about the magnitude of any future violence reduction. That uncertainty belongs in the report. A risk formulation should distinguish demonstrated changes in circumstances, a plausible risk-management benefit, and an empirically supported change in forecast.
An illustrative case makes the logic clear. A person has a documented history of serious violence and recently faced unstable accommodation, problematic drinking, and repeated conflict. Over six months, stable housing and treatment participation improve and incidents decrease. The history remains. Current needs and management conditions may be more favorable. A reviewer can describe those verified changes and specify remaining vulnerabilities, without inventing a new percentage probability or promising that violence cannot recur.
Calibration, Base Rates, and the Meaning of a 'High-Risk' Label
Calibration concerns whether observed outcome rates match the probabilities or categories a tool communicates. A model can rank people better than chance yet systematically overestimate how frequently an outcome occurs. An assessor cannot infer a valid individual percentage from a score unless a suitable calibrated model, target population, and time horizon support that interpretation. Broad verbal categories such as 'low,' 'moderate,' and 'high' require an explicit reference system and can be misleading when used as intuitive adjectives.
Base rates are central. When an outcome is rare, even a seemingly useful classifier can identify many people as elevated risk who will not experience that outcome within the forecast window. Positive predictive value depends on prevalence as well as test sensitivity and specificity. An instrument calibrated in a high-prevalence setting can overstate or understate risk elsewhere. Administrative outcomes such as rearrest are further shaped by surveillance and processing practices.
An illustrative calculation shows the issue without applying it to any person: suppose 100 people are observed and 10 have the outcome. A hypothetical classifier identifies 8 of those 10, but also marks 18 of the 90 people without the outcome as positive. Its sensitivity is 80%, its specificity is 80%, and its positive predictive value is only 8/26, approximately 31%. Twenty-six positive results are not twenty-six future offenses. Real performance depends on the actual tool, cutoff, population, and follow-up window.
False Positives, False Negatives, and Decisions About Liberty
A false positive occurs, in a defined classification exercise, when someone is designated positive for a future event but does not meet the recorded outcome definition during follow-up. A false negative occurs when someone is classified negative and the event is recorded. Each has distinct consequences: unnecessary restrictions and stigma on the one hand, and missed opportunities for prevention on the other. There is no ethical or statistical basis for describing either error as inconsequential.
Observed outcomes can also be changed by the response to the assessment. If intensive treatment, supervision, or an environmental restriction follows a high-risk designation, the absence of a later event might reflect effective management rather than a faulty forecast. Conversely, differences in surveillance can increase detection of technical violations in the closely monitored group. These feedback loops make causal interpretation and evaluation difficult. Ogonah and colleagues discuss the problem for forensic discharge studies.
Evaluators should therefore report the purpose of a threshold, false-positive and false-negative implications where known, uncertainty, subgroup evidence, and available less-restrictive risk-management options. A formal decision about detention, release, sentencing, or other legal consequences belongs to the legally empowered authority applying the relevant local standard. Psychological estimates cannot silently substitute for a jurisdiction's rules or a person's procedural rights.
Cultural Fairness and Population Limits
Risk factors are measured inside social institutions, not in a neutral vacuum. Prior justice contact depends partly on patterns of detection and enforcement; employment and housing opportunities are affected by socioeconomic and structural conditions; interview meaning may differ across language and cultural contexts. Apparent differences in performance across groups can be hidden by an overall AUC. Evaluations require transparent definitions and subgroup checks where feasible, including calibration and the consequences of different error rates.
The APA forensic guidelines direct practitioners to consider individual and group differences and the suitability of instruments for the examined population. The NIJ post-sentencing assessment guidelines describe fairness, effectiveness, efficiency, and communication as important principles. These are professional and research-oriented materials; they do not impose one uniform legal rule across federal, state, or international jurisdictions.
A risk estimate validated among adult men in a specialist hospital may perform differently among women, adolescents, first-time defendants, people with disabilities, or people assessed under another country's legal procedures. Small subgroup samples also create uncertainty about both accuracy and apparent disparities. When validation is missing, the proper statement describes that evidential gap instead of substituting generic claims of objectivity or fairness.
Static and Dynamic Factors in Violence Risk, Recidivism, and Threat Assessment
Violence risk assessment asks a question about interpersonal violence in a particular context. General recidivism assessment asks whether a defined type of justice-system event will recur. Sexual-recidivism instruments focus on a different outcome and often a different reference group. Threat assessment concerns the management of a potential pathway toward targeted harm and may involve behavioral evidence, situational access, and prevention planning. These tasks overlap in practical concerns but cannot share all predictors or operating thresholds without validation.
A previous violent conviction can be historically relevant to one violence assessment while offering little information about an unrelated civil referral. A recent breakdown in community support may matter for management without being a validated standalone indicator of imminent offending. What counts as a factor, how it is rated, and whether it forecasts the intended outcome must follow the assessment context. Generic profiles of 'a dangerous person' collapse these distinctions and create interpretive errors.
What Do Static and Dynamic Factors Mean in Court?
Risk evidence may be presented in bail, sentencing, probation, parole, civil commitment, correctional placement, or clinical-discharge decisions, depending on local legal rules. The statutory question, admissibility requirements, procedures, and evidentiary standards differ substantially. A psychologist can describe the underlying findings, the tool's validity, conditions of assessment, and limits of the opinion. The judge or other authorized decision-maker must apply the legal standard. Risk is not synonymous with guilt, legal insanity, competency, or criminal responsibility.
Risk assessment and competency evaluation also ask different questions. Competency or capacity generally concerns present legally specified abilities; a conviction concerns proof of historical conduct under applicable law; a violence assessment concerns an uncertain future event. Prior offending can be evidence within a structured forecast but cannot determine whether a defendant committed the conduct charged in a new case. A diagnostic label, substance-use history, or psychometric score likewise does not decide culpability.
For a general account of the field and its boundaries, see Forensic Psychology: What It Is and What Forensic Psychologists Do. The distinction between forensic findings and legal conclusions should remain explicit in every report or testimony, especially where liberty, supervision intensity, access to care, or public safety may be affected.
The Relationship Between Mental Illness and Risk Factors
Mental-health diagnosis and future violence are separate matters. Most people living with a mental disorder are not violent, and diagnostic categories are heterogeneous. A diagnosis alone is an inadequate explanation of a criminal act or a valid individual prediction. What matters to a forensic referral is the evidence relating to the stated outcome, the person's actual functioning and circumstances, and the scientific limits of the proposed inference.
Some dynamic conditions intersect with clinical care, including acute distress, substance-related difficulties, or a breakdown in service access. Appropriate care can be important on its own ethical and clinical grounds. An assessment should not imply that every episode of symptoms creates criminal risk or that receiving psychiatric treatment is evidence of dangerousness. A record of treatment may instead document help-seeking and support, and its interpretation requires context.
How Often Should Dynamic Risk Factors Be Reassessed?
Reassessment should respond to the purpose, timescale, setting, instrument guidance, and material changes in circumstances. An acute risk-management concern may require prompt reconsideration, while relatively stable patterns may be reviewed at longer intervals. There is no scientifically defensible universal schedule that applies to every person and outcome. Routine administration can create false precision if ratings fluctuate because different assessors use inconsistent definitions.
Each review should preserve the original baseline, list new information, record changes with dates and sources, and explain what those changes do and do not support. Communication with relevant professionals should be proportionate to the legal authority and confidentiality rules that apply. A report that simply announces 'risk improved' misses the more useful questions of what improved, whether improvement is sustained, what support maintains it, and what uncertainty remains.
A Practical Case Formulation: Integrating History, Change, and Context
Consider a fictional adult nearing the end of a custodial sentence. Historical records show two prior violent incidents years apart. A current review finds reliable participation in vocational work, improved housing prospects, and a recently disrupted family relationship. The evaluator verifies the offense records, distinguishes ordinary relational stress from offending-related warning patterns, and examines whether the selected assessment framework has been validated for the relevant population. No number or classification follows merely from the illustration.
A useful formulation might identify a conditional scenario: if stable housing collapses, contact with certain prior associates resumes, and identified conflict behaviors intensify, the existing management plan may need revision. It would also recognize the person's strengths and the protections available through services and supportive relationships. The formulation makes assumptions and proposed actions inspectable. It does not confuse a list of correlates with the causes of the earlier incidents.
The same case would support distinct outputs for different readers. A treatment team may need individualized service targets; a parole professional may need practical review triggers; a court may need the limits of forecast evidence and the legal relevance of specific findings. Preserving those distinctions is essential to fair use of psychological expertise.
Common Myths and Misconceptions
One misconception is that static factors imply an unchangeable destiny. They preserve historical information, while future behavior remains uncertain and shaped by numerous conditions. A second misconception is that a dynamic factor must be a cause of offending. Its value as a marker and its value as an intervention target require different evidence. A third misconception is that the absence of a major recorded incident proves that an instrument was accurate or inaccurate; active management, opportunity, follow-up length, and data quality complicate that conclusion.
Another misconception is that a high score means the assessor knows what an individual will do. Scores summarize model-defined information and must be interpreted within the validation population, time frame, and intended use. A final misconception is that professional language makes an unsupported judgment more scientific. Technical terms such as 'criminogenic' and 'acute' acquire meaning only through operational definitions, appropriate evidence, and explicit limits.
Questions to Ask About a Forensic Risk Report
A careful reader can ask what adverse outcome is being evaluated, which time horizon applies, and why the selected tool fits the person and setting. Which findings are historical facts and which are current observations? Are they based on independently verified information or a single report? Do risk estimates reflect a properly calibrated model? What is known about false positives, false negatives, and subgroup validity? A well-supported opinion should make those questions answerable.
The report should also distinguish possible management strategies from demonstrated causal effects. If a score has changed, the reader should see what changed in the underlying data, whether the instrument has longitudinal validity, and how the change informs a proportionate plan. These questions protect scientific accuracy and the legitimacy of the legal process.
Frequently Asked Questions
Is age a static or dynamic risk factor?
Age itself changes, while age at first offense or age at a particular historical event remains fixed. Some actuarial instruments update estimates according to current age. The classification depends on exactly which age-related variable is being measured and how the relevant instrument treats it. Older age does not erase a past event, and age trends observed across groups do not dictate an individual's actions.
Is a criminal record a static risk factor?
The record of past incidents or convictions is historical and generally treated as static. Whether the record accurately describes the underlying behavior, and whether its relationship to future outcomes persists in a particular population, are separate questions. Records also differ across jurisdictions in definitions, completeness, expungement rules, and legal consequences.
Is substance use a dynamic risk factor?
Patterns of problematic substance use can change and may be relevant to some tools and outcomes. Mere use or the existence of a substance-use diagnosis is not sufficient for an individual forecast. The evaluator must examine the specified behavior, context, timing, evidence for the relevant outcome, and appropriate treatment needs without assuming that a diagnosis means violence.
What is the difference between stable and acute dynamic factors?
Stable dynamic factors tend to evolve over a relatively extended interval, while acute dynamic factors concern more immediate fluctuations. These are practical categories within particular forensic research traditions; their predictive time scales and measurement quality vary. Acute stress alone is not a demonstrated warning that a person will commit an offense.
Can someone move from high risk to low risk?
A person's circumstances and assessment classification can change, and some instruments allow reclassification. A defensible change in label requires verified new information and evidence that the instrument validly interprets longitudinal changes. The original comparison group, outcome, horizon, and limits remain important. The label is not a legal finding or guarantee.
Do dynamic risk factors predict violence better than static factors?
There is no universal ranking. Relative predictive validity depends on the selected measures, type of violence or recidivism, setting, people assessed, time horizon, and analytic method. Studies sometimes find incremental value for dynamic instruments and sometimes do not. Reviews consistently call for stronger evidence about discrimination, calibration, and change over time.
Do static or dynamic risk factors establish criminal responsibility?
No. Criminal responsibility is evaluated under legal standards set by the relevant jurisdiction and concerns a different question from future risk. Risk factors and assessment scores may appear in particular legal proceedings where admissible, but they cannot replace evidence of the charged event or the applicable legal test.
Can risk factors be used to diagnose someone?
Risk factors, screening results, and forensic score categories are not diagnoses. Diagnosis requires the applicable clinical assessment criteria, and forensic risk assessment requires a separate, explicitly defined psycholegal question. Neither procedure can be inferred from a checklist of behaviors in a newspaper or online profile.
What Makes an Evidence-Based Interpretation?
An evidence-based interpretation identifies the target outcome, population, time horizon, measures, comparison data, uncertainty, and proposed decision context. It distinguishes fixed history from measured change and separates a predictor from a cause. It treats human behavior as probabilistic, ensures that legal findings remain with legal decision-makers, and explains when the available research is insufficient for a numerical or individual claim. The point of the static–dynamic distinction is better specified reasoning, not certainty about the future.
Related Articles
References
American Psychological Association. (2013). Specialty guidelines for forensic psychology. American Psychologist, 68(1), 7–19. DOI: 10.1037/a0029889. Full APA guidelines.
Campbell, D., Cohen, M., Macpherson, G., & O'Rourke, S. (2026). The longitudinal properties of the HCR-20: A systematic review. Aggression and Violent Behavior, 86, 102111. DOI: 10.1016/j.avb.2025.102111.
Fazel, S., Hurton, C., Burghart, M., DeLisi, M., & Yu, R. (2024). An updated evidence synthesis on the Risk-Need-Responsivity (RNR) model: Umbrella review and commentary. Journal of Criminal Justice, 92, 102197. DOI: 10.1016/j.jcrimjus.2024.102197.
Hanson, R. K., & Harris, A. J. R. (1998). Dynamic predictors of sexual recidivism. Department of the Solicitor General Canada. Government research report.
Heffernan, R., Wegerhoff, D., & Ward, T. (2019). Dynamic risk factors: Conceptualization, measurement, and evidence. Aggression and Violent Behavior, 48, 6–16. DOI: 10.1016/j.avb.2019.06.004.
National Institute of Justice. (2011). Re-entry: Dynamic risk assessment (NCJ 238075). U.S. Department of Justice. Official study record.
National Institute of Justice. (2021). Guidelines for post-sentencing risk assessment (NCJ 300654). U.S. Department of Justice. Official report record.
Ogonah, M. G. T., Seyedsalehi, A., Whiting, D., & Fazel, S. (2023). Violence risk assessment instruments in forensic psychiatric populations: A systematic review and meta-analysis. The Lancet Psychiatry, 10(10), 780–789. DOI: 10.1016/S2215-0366(23)00256-0.
Public Safety Canada. (1999). Approaches to offender risk assessment: Static vs dynamic. Corrections research summary. Government research summary.
Public Safety Canada. (2010). Giving meaning to risk factors. Corrections research summary. Government research summary.
van den Berg, J. W., Smid, W., Schepers, K., Wever, E., van Beek, D., Janssen, E., & Gijs, L. (2018). The predictive properties of dynamic sex offender risk assessment instruments: A meta-analysis. Psychological Assessment, 30(2), 179–191. DOI: 10.1037/pas0000454.