Interpreting Data for Referral, Goals, and Service Decisions

Interpreting Within the Strength of the Evidence

After records review, interviews, skill assessment, preference assessment, descriptive assessment, and/or functional analysis, the BCBA decides what the data mean for service. The governing rule: interpret only as strongly as the method allows.

• A confirmed skill deficit → an acquisition goal (teach the missing response).
• A function demonstrated by FA — e.g., behavior maintained by escape from demands → a function-based plan: teach a functional communication replacement (e.g., requesting a break), enrich/modify the demand context, and plan extinction of the problem behavior.
• Data suggesting medical, psychiatric, sensory, speech-language, or legal needs outside scope → refer or collaborate.

A descriptive-only data set does not justify the strong causal language that an FA would. Writing 'behavior is maintained by attention' from correlational ABC data alone overstates the evidence — a classic interpretation error the exam targets.

The right way to hedge: when only descriptive data exist, phrase findings as hypotheses ('the pattern suggests escape; an FA would be needed to confirm') and choose interventions whose risk is acceptable even if the hypothesis is wrong. When an FA has demonstrated function, you may write function-based plans with confidence and use stronger language. Matching the certainty of your language and your intervention to the strength of your method is itself a tested competency.

Prioritizing Goals

When assessment surfaces many possible targets, prioritize systematically rather than by whoever complained loudest.

Safety dominates. Behaviors that risk injury, elopement, or a health crisis are addressed first. After safety, weigh social significance and function together — a goal that is both highly valued by the client and tied to the behavior's function is the strongest target. Feasibility is a real constraint: a perfectly reasoned goal the team cannot implement with integrity will fail in practice, so buy-in and measurability matter.

There is also a sequencing logic between skill-building and reduction goals. Where assessment shows a problem behavior serves a function the learner has no appropriate way to achieve, the replacement skill should be built alongside or before reduction.

The most direct vehicle is functional communication training (FCT), which teaches a communication response that produces the same reinforcer the problem behavior did. Reducing a behavior without giving the learner a functional alternative tends to produce symptom substitution or extinction-induced escalation, which is why function-based replacement — not bare suppression — is the defensible target the exam expects.

Writing Goals and Deciding on Service

Write from data

A goal must specify the learner, behavior (response), condition, criterion, and context. It is not a personality outcome such as 'will be respectful' or 'will have a good attitude' — those are not observable or measurable. Compare:

• Weak: 'Jordan will be more independent.'
• Strong: 'Given a visual schedule and a verbal cue (condition), Jordan will complete the 6-step morning routine independently (response) on 80% of opportunities across 3 consecutive days (criterion) at home and at school (context).'

The strong version is observable, measurable, condition-specified, and tied to the assessment finding.

Service and referral decisions

Assessment may indicate that ABA services are warranted, that a different service is needed first, or that collaboration is required. Example: a communication assessment may support a speech-language referral while the BCBA simultaneously targets functional communication in daily routines. Referral is appropriate whenever needs fall outside the BCBA's competence or behavior-analytic scope.

Interpretation also sets the dosage and intensity of recommended service, not just its content. Severe, high-frequency, dangerous behavior with a confirmed function justifies more intensive, closely supervised programming; a single mild skill gap may need only a brief, targeted plan. The data should defend how much service as well as which goals, and the recommendation should name how progress will be measured so the team can adjust. Recommending more service than the data support, or less than safety demands, are both interpretation failures the exam can probe.

The dead-man's test and operational targets

Reduction goals fail the dead-man's test when they describe what the client will stop or not do — a dead person can 'not be disruptive' perfectly. Strong goals name an active, observable response the living client will do more of. Replace 'will stop eloping' with 'will remain within the activity area and request a break when needed.' This reframing forces a teachable, measurable, function-relevant target and is a frequent exam discriminator between two otherwise similar options.

The exam's favorite trap

Many items embed a tempting goal that is easy for staff but only weakly tied to the referral concern — typically one that suppresses behavior or satisfies the setting's preference (compliance, quietness, stillness). The correct answer is the goal that follows from the assessment data and improves client welfare (access, safety, independence, communication, participation), even when it asks more of the team. When two options both look reasonable, prefer the one that is function-based, socially valid, and passes the dead-man's test — that combination almost always marks the keyed answer.