FASS cheat sheet

Base rate of a disorder sets the upper bound on positive predictive value of any test; low base rates produce many false positives even with high specificity.

Reliability caps validity: a test cannot be more valid than it is reliable (validity coefficient ≤ √reliability).

Standard error of measurement: SEM = SD × √(1 − r); 95% CI ≈ score ± 1.96 × SEM.

IQ scores are normed with mean 100 and SD 15; index scores on WAIS/WISC use the same metric.

A T-score has mean 50 and SD 10; MMPI-2 clinical scale elevations at T ≥ 65 are conventionally clinically significant.

Diagnostic likelihood ratio: LR+ = sensitivity / (1 − specificity); LR− = (1 − sensitivity) / specificity.

Neuropsychological impairment is inferred from discrepancies relative to premorbid estimate and to demographic norms, not from a single low score alone.

Test-retest and interrater reliability coefficients of .90 or higher are generally required for high-stakes individual decisions.

DSM-5-TR criteria for a major depressive episode require ≥5 symptoms for ≥2 weeks, including depressed mood or anhedonia, causing impairment.

Mania requires ≥1 week of elevated/irritable mood plus increased energy; hypomania requires ≥4 days and no marked impairment or psychosis.

Schizophrenia requires ≥2 core symptoms (delusions, hallucinations, disorganized speech, grossly disorganized/catatonic behavior, negative symptoms) with ≥1 from the first three, plus ≥6 months of disturbance.

Antisocial personality disorder requires conduct disorder evidence before age 15 and ≥3 of 7 adult criteria.

Panic attack specifier can apply across disorders; panic disorder requires recurrent unexpected attacks plus ≥1 month of anticipatory worry or behavior change.

Substance use disorder severity: 2–3 criteria mild, 4–5 moderate, ≥6 severe, out of 11 criteria in DSM-5.

Delirium is distinguished from dementia by acute onset, fluctuating attention/awareness, and a reversible medical etiology.

PTSD requires ≥1 intrusion, ≥1 avoidance, ≥2 negative alterations in cognition/mood, and ≥2 arousal/reactivity symptoms lasting >1 month.

Risk assessment tools are actuarial (fixed algorithm), structured professional judgment (SPJ, e.g., HCR-20), or unstructured clinical; actuarial yields better calibrated group estimates.

Violence risk instruments commonly assess historical, clinical, and risk-management (HCR-20) or static and dynamic factors (LSI-R, PCL-R).

PCL-R is not a stand-alone violence predictor; it is a psychopathy measure with factor structure (interpersonal/affective, lifestyle/antisocial).

Sex offender tools (Static-99R, SVR-20) yield relative risk estimates, not individual probability of reoffense for a specific person.

Base rates dominate predictive accuracy; low base-rate populations produce low positive predictive value regardless of instrument.

Malingering assessment uses symptom validity tests (e.g., TOMM, MMPI-2 validity scales) and structured criteria such as Slick et al.

Competency to stand trial evaluates present ability to understand proceedings and assist counsel (Dusky standard); insanity evaluates mental state at the time of the offense.

Risk formulation should distinguish static (unchangeable) from dynamic (modifiable) factors and identify protective factors.

Incompetency restoration typically involves psychotropic medication and education about the legal process; Jackson v. Indiana limits indefinite confinement of unrestorably incompetent defendants.

Sell v. United States sets conditions for involuntary medication solely to restore competency: important governmental interests, medication substantially likely to restore, less intrusive alternatives, and medically appropriate.

Sex offender treatment often uses relapse prevention or Good Lives Model frameworks; risk-need-responsivity (RNR) principles guide intensity matching.

RNR model: risk principle (match intensity to risk), need principle (target criminogenic needs), responsivity principle (match learning style).

Therapeutic jurisprudence examines how legal rules and procedures affect psychological well-being.

Treatment refusal by competent inmates is generally protected; involuntary treatment requires due process (Washington v. Harper).

Diversion and mental health courts aim to reduce recidivism by linking offenders to community treatment with judicial monitoring.

Outpatient commitment statutes vary by jurisdiction; empirical support for reduced recidivism is mixed and depends on service availability.

Informed consent requires capacity, disclosure of risks/benefits/alternatives, voluntariness, and documentation.

Confidentiality is limited by duty to protect (Tarasoff), mandated reporting of child/elder abuse, and court order.

Multiple relationships are discouraged when they could impair objectivity or exploit; not all are unethical, but risk must be managed.

Test data vs. test materials: APA Ethics Code distinguishes release of raw data (client rights) from protection of test security.

Privilege belongs to the client and can be waived; forensic evaluations generally carry limited confidentiality because the court is the client.

HIPAA permits disclosure for treatment, payment, and operations; psychotherapy notes require separate authorization.

Ethics codes (APA) address competence, informed consent, privacy, and avoiding harm; enforcement is by licensing boards and ethics committees.

Record retention and release follow state law, HIPAA, and professional guidelines; minimum retention often 7 years or until age of majority plus statute.

Dusky v. United States (1960) standard: defendant must have sufficient present ability to consult with counsel with a reasonable degree of rational understanding and a rational/factual understanding of proceedings.

M'Naghten test: defendant did not know the nature/quality of the act or did not know it was wrong due to mental disease.

ALI/MPC insanity test: lack of substantial capacity to appreciate wrongfulness or conform conduct to law.

Guilty but mentally ill (GBMI) verdicts allow conviction with mental illness finding; defendant still punished with treatment.

Competency is a present-state issue and can be raised at any time; insanity addresses state at offense and is an affirmative defense.

Miranda waiver must be knowing, intelligent, and voluntary; custody plus interrogation triggers warnings.

Diminished capacity is not a full defense in most jurisdictions but may negate specific intent mens rea.

Automatism and unconsciousness are complete defenses distinct from insanity; burden allocation varies by jurisdiction.

Civil commitment requires mental illness plus danger to self/others or grave disability, with due process (Addington v. Texas: clear and convincing standard).

Guardianship/conservatorship involves determination of incapacity and best interests; lesser restrictive alternatives should be considered.

Personal injury claims require duty, breach, causation, and damages; psychological damages must be causally linked to the event.

Workers' compensation and disability evaluations use functional capacity and impairment ratings (e.g., AMA Guides).

Fitness-for-duty and disability determinations differ from clinical diagnosis; they assess functional ability to perform job requirements.

Child custody evaluations apply best-interests-of-the-child standard, considering stability, parental capacity, and child needs.

Testamentary capacity requires knowing the nature of the act, the property, natural heirs, and the plan of distribution.

Civil competency standards (contracts, wills) are generally lower than criminal competency standards.

Expert testimony admissibility follows Daubert (federal) factors: testability, peer review, error rate, general acceptance; Frye uses general acceptance only.

Forensic experts should distinguish treating clinician role (advocacy for patient) from forensic evaluator role (objective to the court).

Opinion testimony must rest on sufficient facts/data, reliable principles/methods, and reliable application (Federal Rule of Evidence 702).

Ultimate issue testimony may be permitted but should not usurp the trier of fact; many jurisdictions limit it.

Report writing should separate observations, data, inferences, and opinions; include limitations and alternative hypotheses.

Consultation models include client-centered, consultee-centered, and program-centered administrative consultation (Caplan).

Depositions and cross-examination require distinguishing facts known, data reviewed, and opinions formed; concede limits honestly.

Fee arrangements and referral sources must be disclosed; contingent fees in forensic work create bias and are generally avoided.