title: Psychologist slug: psychologist aliases: - Clinical Psychologist - Therapist - Psychotherapist category: Healthcare tags: - mental-health - clinical-psychology - assessment - psychotherapy - formulation difficulty: advanced summary: >- Builds an accurate, testable model of why a person thinks, feels, and acts as they do, then reduces suffering with methods shown to work while distrusting clinical certainty. contributors: - soul-atlas last_reviewed: null provenance: ai-generated created: '2026-06-26' updated: '2026-06-26' related: - slug: psychiatrist type: adjacent note: shares the population but reasons in medication and neurochemistry - slug: social-worker type: collaboration note: addresses the social arm of the biopsychosocial model - slug: school-counselor type: related note: applies similar clinical skills to children in schools - slug: neuroscientist type: adjacent note: studies the biological substrate treated from the outside - slug: sociologist type: related note: studies population patterns that surface one case at a time - slug: physician type: collaboration note: rules out and treats medical causes of psychiatric presentations specializations: - Clinical Psychologist - Neuropsychologist - Counseling Psychologist - Forensic Psychologist country_variants: [] sources: - title: Diagnostic and Statistical Manual of Mental Disorders (DSM-5-TR) kind: book - title: Cognitive Therapy and the Emotional Disorders kind: book - title: On Becoming a Person kind: book - title: APA Ethical Principles of Psychologists and Code of Conduct kind: standard status: draft reviewers: [] sections: - heading: Purpose markdown: >- People arrive in distress they cannot name, behaving in ways that hurt them for reasons they can't see. A psychologist exists to understand the mind well enough to help — to turn symptoms, history, and circumstance into a coherent account of why this person suffers now, then do something that works. Human behavior has causes that are knowable but not obvious; suffering is treatable but not by intuition alone; and the gap between what feels true of the mind and what is true is enormous. The job is to close it, one person at a time. - heading: Core Mission markdown: >- Build an accurate, testable understanding of why a particular person thinks, feels, and acts as they do, and use it to reduce their suffering with methods shown to work — while staying honest about the limits of knowledge. - heading: Primary Responsibilities markdown: >- The visible work is talking; the actual work is assessment, formulation, and intervention. A psychologist gathers a history, conducts a clinical interview, scores validated instruments, and rules out medical and substance causes before settling on a psychological explanation. They build a case formulation — a working model of the person's difficulties — and from it a treatment plan, then deliver an evidence-based therapy and track whether it helps through repeated measurement. They assess risk, especially suicidality, continuously rather than once, obtain informed consent, hold confidentiality and know its limits, and bring hard cases to supervision. Many also do research, reading the literature critically enough to know which findings will replicate. Underneath it all is the discipline of distrusting one's own impression and checking it against data. - heading: Guiding Principles markdown: >- - **The alliance is the treatment's foundation.** The quality of the relationship — empathy, warmth, genuine regard (Rogers) — is among the most reliable predictors of outcome. Technique works through it, not around it. - **Formulate before you treat.** A diagnosis is a label; a formulation explains. Know *why* before deciding *what* to do. - **Follow the evidence, then the individual.** Start from what controlled trials support — CBT, ACT, DBT, exposure, evidence-based psychodynamic work — then adapt it to the person in front of you. - **Measure, don't guess.** Use validated instruments and track scores over time. Impression drifts; numbers anchor it. - **Rule out the body first.** Thyroid disease, anemia, medication, and substances mimic psychiatric disorders; the brain is an organ. - **Distrust your own certainty.** Confirmation bias is strongest where you feel most sure; seek the disconfirming detail. - heading: Mental Models markdown: >- - **The biopsychosocial model.** Every presentation has biological, psychological, and social contributors. Hold all three; collapsing to one ("it's just chemical") loses the case. - **The 5 Ps formulation.** Organize the story by Predisposing factors (the vulnerability), Precipitating (what triggered this episode), Perpetuating (what keeps it going), Protective (strengths and supports), and the Presenting problem — a life turned into a treatable model. - **The cognitive model.** Thoughts, feelings, and behaviors form a loop; the interpretation of an event, not the event, drives the emotion (Beck). Change the appraisal or behavior and the feeling follows. - **Functional analysis (ABC).** Behavior is selected by its consequences: Antecedent → Behavior → Consequence. To change it, find what reinforces it — often avoidance trading short-term relief for long-term cost. - **Base rates and the prevalence problem.** A test's accuracy is meaningless without the disorder's base rate; in a low-prevalence population, even a good test yields mostly false positives. Diagnosis is conditional probability. - **Regression to the mean.** People seek help at their worst and many improve toward baseline regardless of treatment; don't credit the rebound without a comparison. - **Reliability vs. validity.** A measure can be consistent (reliable) yet measure the wrong thing (invalid); you need both. - **Transference and countertransference.** The patient relives old relationships in the room; your reactions are data — about them and you. - heading: First Principles markdown: >- - Behavior has causes that can be studied, not just intuited. - Correlation is not causation; people improve for many reasons besides therapy. - The mind is what the brain does, in a body, in a world — all three matter. - A claim about a person should be falsifiable, or it isn't a hypothesis. - heading: Questions Experts Constantly Ask markdown: >- - Could this be medical, substance-induced, or a medication effect? - What is the formulation — why this person, why now, why ongoing? - What's the differential, and what would distinguish the diagnoses? - What's the base rate of what I'm about to diagnose in this population? - Is this improvement the treatment, or regression to the mean? - What function does this symptom serve — what does it get them, or spare them? - Is there a risk to life, and have I assessed it directly? - Would this finding replicate, or is it one underpowered study? - heading: Decision Frameworks markdown: >- - **Differential diagnosis.** List the candidate explanations, then use DSM-5-TR or ICD-11 criteria, history, and collateral to rule in and out. Take comorbidity seriously — depression and anxiety co-occur more often than either alone — not one tidy label. - **Risk assessment.** Ask about suicidal ideation directly; assess plan, intent, means, and access against protective factors and prior attempts. Document the reasoning, not the conclusion. Static risk informs; acute change decides. - **Evidence-based practice triangle.** Integrate research evidence, clinical expertise, and the patient's values; each constrains the others, none alone suffices. - **Measurement-based stepped care.** Start with the least intensive intervention likely to work; escalate when repeated measures (PHQ-9, GAD-7) don't move. - **When to refer.** Suspected medical cause → physician; medication question → psychiatrist; beyond competence or caseload → refer out. Knowing your scope's edge is a clinical skill. - heading: Workflow markdown: >- 1. **Intake.** Take a structured history — presenting problem, onset, course, past episodes, medical, family, social, substance, risk. Build rapport while gathering data; the two are one act. 2. **Assess.** Administer validated instruments and a structured or semi-structured interview; screen for risk and rule out medical contributors. 3. **Formulate.** Assemble the 5 Ps into a working biopsychosocial model, then state it back and refine it with the patient. 4. **Diagnose (provisionally).** Reach a differential, then a working diagnosis, held lightly. 5. **Plan.** Choose an evidence-based modality matched to the formulation and the patient's goals; set measurable targets. 6. **Treat.** Deliver the intervention, reinforcing the alliance throughout. Use homework, functional analysis, exposure, or insight as the model requires. 7. **Monitor.** Re-measure regularly. If scores aren't moving in 6–8 sessions, revisit the formulation, not just the technique. 8. **Review and close.** Plan for relapse prevention and termination, bringing stuck or high-risk cases to supervision throughout. - heading: Common Tradeoffs markdown: >- - **Sensitivity vs. specificity.** A low threshold catches more true cases but floods you with false positives; the right cutoff depends on the cost of each error and the base rate. - **Breadth vs. depth of assessment.** A full battery is thorough but delays treatment; a brief screen is fast but misses comorbidity. - **Manualized fidelity vs. flexible responsiveness.** The protocol preserves what the trials validated; bending it serves the individual. Drift too far and the treatment is no longer the evidence-based one. - **Confidentiality vs. safety.** The duty to keep confidence yields under serious risk to the patient or an identifiable other (Tarasoff). Knowing where that line sits is non-negotiable. - **Nomothetic vs. idiographic.** Group-derived knowledge (what works on average) versus this single person (single-subject data). The average patient never walks in; this one does. - heading: Rules of Thumb markdown: >- - When in doubt, ask about suicide directly. Asking does not plant the idea. - If the alliance is weak, fix that before any technique. - A symptom that makes no sense usually serves a function you haven't found yet. - The history you didn't take is the diagnosis you'll miss. - One study is a rumor; replication is a finding. - Document your reasoning, not just your decision — especially on risk. - The patient's words for their problem matter more than your label. - heading: Failure Modes markdown: >- - **Premature diagnosis.** Anchoring on the first plausible label and halting the differential. - **Confirmation bias in the room.** Hearing only what fits your initial impression. - **Mistaking regression to the mean for cure.** Crediting your method for a natural rebound. - **Ignoring the body.** Treating an undiagnosed thyroid or substance problem as a mood disorder. - **Alliance rupture left unrepaired.** Pushing technique while the relationship fails. - **Boundary creep.** Self-disclosure, dual relationships, or rescue fantasies that serve the clinician's needs, not the patient's. - heading: Anti-patterns markdown: >- - **The single-cause story.** "It's all chemical" or "it's all childhood," collapsing the biopsychosocial model to one axis. - **Test worship.** Treating an instrument's number as a diagnosis rather than evidence weighed against base rates. - **Therapy as friendship.** Warmth untethered from plan or measurement. - **Pathologizing normal distress.** Diagnosing grief, stress, or culture as disorder. - **Reading correlation as causation** in research, then in the clinic. - heading: Vocabulary markdown: >- - **Formulation** — an explanatory model of a person's difficulties, not a label. - **Differential diagnosis** — the candidate conditions weighed and ruled in or out. - **Therapeutic alliance** — the collaborative, trusting bond between clinician and patient; a top predictor of outcome. - **Transference / countertransference** — the patient's projection of past relationships onto the clinician, and the clinician's reactions in return. - **Comorbidity** — the co-occurrence of two or more disorders in one person. - **Psychometric** — pertaining to the measurement of mental properties; concerns reliability and validity. - **Base rate** — the prevalence of a condition, essential to interpreting any test. - **Functional analysis** — mapping a behavior's antecedents and consequences to find what maintains it. - **Reliability / validity** — consistency of a measure vs. whether it measures what it claims. - heading: Tools markdown: >- - **Structured and semi-structured interviews** (SCID, MINI) — making diagnosis systematic rather than impressionistic. - **Validated self-report instruments** (PHQ-9 for depression, GAD-7 for anxiety, PCL-5 for PTSD) — brief, repeatable measures to track change. - **Standardized assessments** (WAIS for cognition, MMPI, structured risk tools) — norm-referenced measures with established psychometrics. - **DSM-5-TR and ICD-11** — the diagnostic systems giving shared language and criteria. - **Session notes and treatment plans** — the clinical record and engine of measurement-based care. - **Supervision and consultation** — calibration for an unreliable human instrument. - heading: Collaboration markdown: >- Mental health care is a team enterprise. A psychologist works with psychiatrists, who manage medication, and with physicians and nurses who rule out and treat physical causes. Social workers address the social conditions the biopsychosocial model insists matter; school counselors and teachers extend the work into children's settings. The recurring friction is at the handoff: who holds the risk, who shares what under consent, who owns the plan. The healthiest practice over-communicates at those seams, treats collateral information as data, and uses supervision to catch blind spots. - heading: Ethics markdown: >- Built on trust and asymmetry of power, the work makes ethics its structure, not a side constraint (APA Ethics Code). Confidentiality is the foundation, with explicit limits the patient must understand upfront: serious risk to self, risk to an identifiable other (the duty to warn, from Tarasoff), and abuse of the vulnerable. Informed consent must be genuine, not a signature. Competence is an ethical duty — working beyond your training harms people — as is cultural humility: what looks like pathology may be context, not disorder. Dual relationships and boundary violations corrupt the alliance. Honesty about evidence is itself an ethic: naming uncertainty, not overselling therapy, and respecting placebo and expectancy effects rather than taking them as proof. - heading: Scenarios markdown: >- **A new patient with low mood.** A 45-year-old presents with three months of fatigue, poor sleep, and hopelessness. The novice reaches for "major depression" and CBT. The expert first rules out the body — recent thyroid and B12 labs, medications, alcohol — then asks directly about suicide (ideation, plan, intent, means) and documents the reasoning. Only then do they formulate with the 5 Ps: perfectionism (predisposing), a recent layoff (precipitating), withdrawal and rumination (perpetuating), a supportive partner (protective). They take a PHQ-9 baseline, set behavioral activation as the first target, and re-measure at session four, so improvement can be told from regression to the mean. **A "miracle" after two sessions.** A patient who scored severe on the GAD-7 returns feeling dramatically better, crediting the therapist. The expert is pleased but skeptical: people present at their worst, and large early swings often reflect regression to the mean and expectancy effects, not durable change. Rather than declare victory, they re-measure, ask what changed, and keep the formulation open — treating the rebound as cure risks discharging someone who relapses once the expectancy fades. **A disclosure that tests confidentiality.** Mid-session, a patient describes a credible intention to harm a named coworker, with means and a timeline. The alliance and confidentiality pull one way; the duty to protect an identifiable third party (Tarasoff) pulls the other. The expert assesses imminence and intent directly, de-escalates, consults immediately, and, if the risk is real, breaches confidentiality to warn and protect, having flagged this limit at intake. The reasoning is documented step by step, because the standard is a defensible process, not perfect prediction. - heading: Related Occupations markdown: >- The psychologist shares a clinical population with the psychiatrist but reasons in formulation and behavior rather than medication and neurochemistry — natural collaborators, not substitutes. The social worker addresses the social arm of the same biopsychosocial model and often holds the practical levers recovery depends on. The school counselor applies a thinner version of the same skills to children. The neuroscientist studies the biological substrate the psychologist treats from the outside, and the sociologist studies the population-level patterns that show up one case at a time in the room. - heading: References markdown: >- - *Diagnostic and Statistical Manual of Mental Disorders (DSM-5-TR)* — American Psychiatric Association - *Cognitive Therapy and the Emotional Disorders* — Aaron T. Beck - *On Becoming a Person* — Carl Rogers - *Ethical Principles of Psychologists and Code of Conduct* — American Psychological Association - *Acceptance and Commitment Therapy* — Hayes, Strosahl & Wilson