MFT National Exam common mistakes
MFT National Exam · Understand the traps and how to avoid them.
Reads a vignette describing a couple's or family's escalating circular pattern and selects an answer aimed at the identified client's individual coping skills instead of at the interactional pattern.
Most trainees were trained in individually oriented models and default to symptom-reduction language; the exam consistently rewards a systemic frame, so plausible-sounding individual answers are the strongest distractors.
What to do instead
Before looking at the options, write one sentence naming the circular pattern (who does what, in response to what). Eliminate any option that treats one person's behavior as the cause or as the only treatment target.
Attributes a technique to the wrong model - for example calling a structural enactment 'paradoxical intention', or crediting solution-focused therapy with genogram work.
Techniques across models overlap in everyday language, so under time pressure candidates match on the surface word rather than on the model's core assumptions.
What to do instead
Build a one-page grid: model, key theorist, signature technique, view of the problem. Drill until you can reject an option because the technique contradicts that model's epistemology (directive/strategic vs. collaborative vs. experiential).
Selects a DSM diagnosis on the first pass without first eliminating medical, substance, or medication-induced explanations for the presenting symptoms.
Vignettes foreground dramatic symptoms and a diagnostic label feels like the 'real' answer, while rule-out language looks like an evasion of the question.
What to do instead
Run a fixed three-step screen on every diagnostic item: medical condition, substance or medication, psychosocial stressor. When those are unresolved, the correct option typically involves screening, referral, or deferring the diagnosis.
Assigns an individual disorder when the stem's data better support a relational problem or psychosocial stressor, or misses that a relational issue and an individual disorder are both present.
Candidates equate 'diagnosis' with 'individual DSM disorder' and forget relational and contextual codes and the possibility of co-occurring conditions.
What to do instead
Ask whose problem this is and whether the distress is explained by the relationship or context. Practice sorting the same vignette into relational problem, stressor, individual disorder, and dual diagnosis, and check whether the correct answer requires more than one.
Answers a 'what should the therapist do FIRST/NEXT' stem with the definitive long-term intervention rather than the immediate step such as joining, safety, informed consent, or assessment.
The most sophisticated clinical answer is often the most attractive choice, but these items test sequencing rather than depth of intervention.
What to do instead
Underline the sequencing word in the stem. Rank the options by time horizon and choose the earliest one that is clinically defensible; discard any option that requires information the therapist does not yet have.
Chooses a powerful experiential technique - enactment, revealing one partner's secret, confrontation - in a vignette showing intimate partner violence, active substance use, or acute crisis.
Technique knowledge is retrieved without checking indications and contraindications, so the option simply 'sounds like good therapy'.
What to do instead
Before selecting a technique, list the contraindications present in the stem. If safety, consent, or stabilization is unresolved, the correct answer is containment, safety planning, or referral rather than the technique.
Ends treatment because symptoms have improved, without addressing goal attainment, generalization of gains, relapse prevention, or a planned termination process.
Symptom reduction is the most salient cue in a vignette and feels like a natural endpoint, while termination skills receive less rehearsal than intervention skills.
What to do instead
Check the stem against the original treatment contract: were the goals met, have skills been transferred to the client's system, and is there a plan for setbacks and the next transition? Correct options usually include reviewing progress, anticipating relapse, and offering step-down or booster sessions.
In couple or family work, reveals one member's individual disclosure to the other, or agrees in advance to keep secrets without ever clarifying the limits of that agreement.
Habits from individual-therapy confidentiality carry over, and the vignette creates emotional pressure to 'be honest' or to 'respect privacy' without a policy.
What to do instead
State the policy at the outset: who the client is, what will be shared, and that no secret will be kept that undermines the treatment. On items, the answer that clarifies the policy before any disclosure beats both the immediate reveal and the flat refusal.
Misses the mandated-reporting or duty-to-protect judgment - either waiting for proof before reporting suspected child abuse, or assuming a warning is automatically required for any angry statement a client makes.
Candidates import a 'certainty' standard into reporting and treat Tarasoff as a single national rule rather than a state-specific duty.
What to do instead
Memorize the thresholds: report on reasonable suspicion, not proof; duty to protect is state-law dependent and is triggered by an identifiable victim plus a serious threat, with options that may include warning the victim, notifying police, or arranging hospitalization. With an identifiable victim and serious threat, act; with diffuse anger, do not.
Confuses research concepts in vignette items - treating reliability as validity, reading a correlation as evidence of causation, or equating evidence-based practice with using a manualized model.
The terms are memorized as vocabulary pairs rather than as distinct claims, and exam items exploit the familiar-sounding pairing.
What to do instead
Drill each pair with a one-line definition and an example: reliability is consistency, validity is measuring what is claimed, sensitivity catches true cases, specificity rules out non-cases. Then ask of each study item: what claim can this design actually support?