ABPP Clinical Child & Adolescent common mistakes
ABPP Clinical Child & Adolescent · Understand the traps and how to avoid them.
Applying adult-derived symptom thresholds to young children - labelling a 2-year-old's tantrums as oppositional defiant disorder, a preschooler's normal fear of dogs as a specific phobia, or a 4-year-old's imaginary companions as psychosis - and selecting that label as the answer.
Most training is adult-heavy, so the vivid vignette behaviour matches an adult template; candidates also forget that frequency, intensity, and pervasiveness must be judged against the normative trajectory for that exact age.
What to do instead
Memorise age-of-onset and normative-timeline tables for the high-yield disorders, and before answering ask three questions: is this behaviour expected at this age, is there impairment across settings, and does it deviate from the developmental expectation rather than merely from adult norms?
Anchoring on the referral question: the stem says the school suspects ADHD, and the candidate picks the ADHD treatment option without first generating alternatives such as anxiety, trauma exposure, sleep-disordered breathing, hearing loss, absence seizures, giftedness with boredom, or a learning disorder.
Referral bias is powerful and the vignette supplies confirmatory details; under time pressure the first plausible formulation feels like a differential diagnosis.
What to do instead
Force a written three-item rule-out list before reading the options - one medical, one psychosocial, one developmental or educational mimic - and eliminate any option that treats a diagnosis that has not survived the rule-out; practice stems that deliberately bury a mimic.
Instrument misuse: picking a measure because the name is familiar even though it is outside its validated age range or norm group, diagnosing from a single parent-completed rating scale, or treating a screening result as diagnostic, including interpreting one informant's score as the child's true level.
Candidates study constructs rather than test characteristics, and the options often include a well-known instrument that is age-inappropriate for the vignette; informant discrepancy is also poorly understood.
What to do instead
Build a one-page matrix of two or three measures per high-yield construct with age range, informant, norming sample, and psychometric properties; apply the multi-informant, multi-method rule, interpreting cross-informant discrepancies as data about situational specificity rather than as error to be averaged away.
Choosing the intuitively appealing or systemic treatment over the age-appropriate empirically supported first line - for example, individual CBT for a 4-year-old with disruptive behaviour instead of parent management training, or insight-oriented play therapy for paediatric OCD instead of exposure and response prevention.
Answers are written to sound humane and family-centred, and candidates reason from general theory rather than from the disorder-by-age evidence base.
What to do instead
Memorise a disorder-by-age-band grid of first-line treatments, including the parent-mediated and school-mediated variants, and rehearse the question 'what is the best-supported treatment for this specific presentation at this age?' rather than 'what would a good clinician do?'
Recommending an unmodified protocol without addressing moderators - a standard 12-session group programme for a child with intellectual disability, limited English proficiency, or comorbid autism - and abandoning the evidence base entirely after one failed trial instead of stepping up or adapting.
Candidates learn that a treatment is 'empirically supported' as a single fact and do not study for whom, under what conditions, and with what adaptations it works, so they treat fidelity and flexibility as opposites.
What to do instead
For each first-line treatment, learn its known moderators, the documented adaptations and their evidence, and the stepped-care logic including measurement-based monitoring; when the stem describes a treatment failure, choose re-assessment and a change in dose, modality, or target rather than passive continuation.
Treating consultation as one-way advice-giving: writing 'refer to psychiatry' or 'tell the school to evaluate' as the final answer, without specifying consent, role clarification, information-sharing limits, or closed-loop communication, and misreading school-system questions such as eligibility under special education law versus a accommodation plan.
Consultation is the least taught domain in clinical training, and vignettes reward a decisive-sounding recommendation, so candidates skip the process steps that the items are actually testing.
What to do instead
Use a fixed consultation template - define the presenting problem, clarify who the client is and what consent covers, state each party's role and constraints, agree on a communication loop, and specify what counts as follow-up; separately learn the differences among school eligibility categories, accommodation plans, and records-privacy rules.
Misapplying ethics rules in either direction: refusing to disclose any information when a child is at risk, releasing a full record to a non-custodial parent, seeing a 15-year-old for ongoing therapy without addressing consent and confidentiality limits, or accepting a court order as automatically overriding clinical judgment.
Candidates either apply the single rule they remember most vividly - usually confidentiality - or default to the most cautious-sounding option, without working through who the client is, what the legal status of each caregiver is, and which exception actually applies.
What to do instead
Use a structured decision model for every ethics stem: identify the client and consent holders, state the governing duty, list the exceptions, consider the least intrusive effective action, and document the rationale; drill custody, divorce, and adolescent consent scenarios until the legal status of each party is automatic.
Misattributing culture to pathology or pathology to culture: reading a child's silent deference, bilingual language mixing, spiritual experience, or culturally normative separation distress as a deficit, or dismissing genuine impairment as 'just cultural', and using the child or a sibling as the interpreter.
Candidates rely on their own cultural defaults when the vignette supplies no explicit cue, and interpreter logistics are rarely practised, so the shortcut of a family member interpreting feels efficient.
What to do instead
Apply a cultural formulation to every vignette - identity, explanatory model, stressors and supports, the clinician-child-family relationship - verify whether norms, translations, and normative data fit the family's language and culture, and always use a trained interpreter with a pre-session briefing, never a family member.
Research-methods and outcome-evaluation confusions: swapping mediator and moderator, treating a statistically significant result as clinically meaningful, over-generalising a single efficacy trial to a dissimilar clinic population, and ignoring how low base rates change the meaning of a positive screening result.
Methodology is studied as vocabulary rather than as reasoning, so similarly named terms collapse under time pressure, and candidates answer from the p-value rather than from the size and applicability of the effect.
What to do instead
Drill worked examples with numbers, especially base-rate arithmetic for a positive screen in a low-prevalence population, and practise classifying variables as moderator, mediator, or confounder in published-style abstracts until the distinction is automatic; always ask how well the sample resembles the child in the stem.
Answering supervision items as generic management: directing a supervisee's case without establishing informed consent for supervision, failing to give documented formative feedback, or responding to a struggling trainee with immediate removal rather than a due-process sequence of notice, remediation plan, and review.
Supervision is often learned on the job rather than studied, so candidates import workplace instincts and forget that supervision itself is an intervention with legal and ethical obligations, including vicarious liability for the supervisor.
What to do instead
Learn supervision as a competency framework - informed consent with the client and the supervisee, a written supervision contract, direct observation, scheduled evaluation, and a documented remediation sequence with appeal rights - and check each option against the question 'does this protect the client, the trainee, and the public?'