How to Ask for Plain-Language Explanations of ABA Data

Start with the exact question

Ask whether the graph is meant to show baseline, progress, maintenance, generalization, treatment integrity, caregiver implementation, access, attendance, or another outcome. The same line can mean little without its purpose.

The CDC service-access page offers general service-navigation information. It does not prescribe ABA measurement methods.

Define the unit and opportunity

Ask what counts once, what starts and stops the measure, and which opportunities are eligible. For a percentage, request the numerator and denominator. For duration, request the start and end events. For rate, request the observation time.

Ask how prompts, AAC, ordinary supports, assent withdrawal, invalid opportunities, staff absence, and environmental failures are recorded. Definitions should match the child's accessible communication and the actual setting.

Read the graph in context

Request the dates, axis labels, condition changes, goal revisions, staff or setting changes, and gaps in collection. A rising or falling line can coincide with several changes.

The CASP ABA Practice Guidelines Version 3.0 public summary places evaluation within individualized ABA behavioral health treatment. It does not supply one universal graph or mastery rule.

Ask for counts beside percentages

“80%” could mean 4 of 5, 8 of 10, or 80 of 100. Those samples carry different uncertainty and burden. Ask whether pending, missed, or excluded observations remain visible.

Plain-language explanations of ABA data should name the cohort and period before presenting a percentage. A polished dashboard cannot repair an unclear denominator.

Include the child's experience

Ask how the person reports comfort, effort, usefulness, preference, or burden. Keep communication available during review. The ASHA AAC portal supports continuous AAC access.

Progress on a narrow measure may coexist with poor fit elsewhere. Ask about school, sleep, relationships, daily life, family priorities, unwanted effects, and quality of life when relevant.

A fictional data question

Nico's graph shows 75% independent requests. The family asks for counts and learns the result is 6 of 8 eligible opportunities across two sessions. AAC was available in both sessions, and one other session lacks data because the device backup failed.

The family asks to keep the missing session visible, review the access failure, and avoid treating 75% as a stable long-term result. The percentage describes the eight eligible opportunities only.

Ask what decision follows

The BACB Ethics Code addresses assessment, intervention, data, documentation, client involvement, risk, and evaluation for covered certificants and applicants.

Ask the responsible clinician what options are being considered, which evidence supports each, how the child and family will participate, and when the team will review the result. Record uncertainty and alternate explanations.

Keep a small data glossary

Save the current definition, unit, calculation, data source, graph date, and clinical contact for measures the family sees often. Update it when the plan changes.

Bring one or two focused questions to each review. Understanding the measure helps the family discuss fit and next steps without turning caregivers into data analysts.

Use a graph question card

Ask: What is measured? What counts once? What is the time period? What are the numerator and denominator? Which sessions or opportunities are missing? Which prompts and ordinary supports were present? What changed in the plan, people, setting, or definition?

Then ask the clinician to summarize the pattern in one sentence and identify the next decision. Write the answer beside the graph date. If the definition changed, request a visible marker so the earlier and later data are not presented as one unchanged measure.

Notice explanation gaps

Pause when a percentage lacks counts, a graph lacks dates, an average hides a wide range, a trend ignores missing sessions, or a conclusion exceeds the observation. Ask for the underlying record or a clearer explanation through the appropriate access process.

Also ask about unwanted effects and outcomes the graph omits. A technically accurate measure can still be too narrow for the family question. The child and family can request a measure that better reflects a meaningful daily-life outcome.

Ask for one worked example

Choose a single date or session and ask the clinician to show how the raw observation became the displayed value. Follow the count, opportunity rule, prompt status, exclusion, calculation, and graph point. A worked example often reveals which part needs clarification. Ask whether the same rule was used across every point and how corrections appear when the record changes.