How to Read an ABA Progress Report: Goals, Graphs and Questions to Ask

First, identify what the report covers

A progress report should tell you whose care is described, which dates are included, who prepared the report, and what decision the report supports. Look for the reporting period, service settings, responsible clinician, relevant diagnoses or referral context, treatment-plan date, and date of the next review.

The report may serve several jobs. A family-facing clinical review explains what the team has learned and what should happen next. A payer reauthorization packet may add plan-specific fields, service codes, requested units, signatures, outcome measures, and medical-necessity rationale. A school report may follow educational requirements. One document can support several jobs when each purpose remains clear.

This walkthrough explains how to read an ABA progress report used for clinical and family review. Apply any payer or school fields according to the governing plan, program, state, and setting.

The CDC autism resource center describes wide variation in autistic people's abilities and needs. Read every finding as information about this child during the stated period and conditions. A group average, diagnostic label, or another child's trajectory cannot supply the missing context.

Read each goal as a complete measurement statement

A goal becomes interpretable when the reader can tell what will happen, under which conditions, with what support, and how success is counted. Find these parts before looking at the latest percentage.

Words such as “appropriate,” “calm,” “cooperative,” “understands,” or “improved behavior” need operational detail. Ask what was actually observed. A family should also be able to connect the goal to communication, access, safety, daily participation, relationships, or another priority they recognize.

The public summary of the CASP ABA Practice Guidelines says the guidelines address planning, implementing, and evaluating assessment and treatment services. The full Version 3.0 document is licensed. This article uses the public summary and does not reproduce a proprietary report form or guideline text.

Use the baseline as the comparison point

Baseline describes performance before a treatment phase or material plan change. It should use the same definition and a comparable measurement system whenever possible. Check how many observations contributed, where they occurred, who collected them, which supports were present, and whether the opportunities resemble current life.

A baseline of “0%” from one opportunity carries less information than 0 independent responses across 12 opportunities in three routines. A baseline collected only in a quiet clinic may poorly represent a goal intended for busy community settings. When the team changes the definition, prompt level, measurement method, or opportunity type, the report should mark a new phase and explain which comparisons remain valid.

Read a graph in a fixed order

A graph can make change visible, though it still needs labels and context. Use the same order each time:

1. Read the title. Confirm the goal, response, setting, and measurement shown.
2. Check both axes. The horizontal axis usually shows sessions, dates, or weeks. The vertical axis may show count, percentage, duration, rate, latency, or another measure.
3. Inspect the scale. A vertical axis from 80% to 100% makes small changes look large. An axis from 0% to 100% shows the full range.
4. Find phase-change lines and event annotations. Baseline and treatment changes should be marked. Prompt changes, medical events, school breaks, and revised definitions should have dates and labels.
5. Look at level. Estimate the typical height or range within a phase rather than focusing on the best point.
6. Look at trend. Describe the overall direction and how long it has lasted.
7. Look at variability. Large swings may reflect different settings, people, opportunity types, health, sleep, measurement, or unstable performance.
8. Find missing data. Blank dates, excluded opportunities, cancellations, and invalid observations should be explained.

The current BACB Ethics Code applies to BCBA and BCaBA certificants and applicants. Standards 2.17 and 2.18 require covered behavior analysts to select and correctly implement data procedures, graphically display or summarize data for decisions, continually evaluate interventions, and take corrective action when desired outcomes are not being realized. The BCBA Test Content Outline is an examination blueprint rather than a reporting standard; it separately identifies measurement validity, representative data, graph interpretation, procedural integrity, experimental design, and data-based treatment decisions as professional knowledge areas.

A percentage needs its numerator, denominator, and support level

Percentages can conceal important differences. This fictional table shows four weeks of an independent help-request goal. Each row uses invented data and represents no actual child.

In this example, “independent” means no trial-specific adult model or gesture; the visual cue remains available as an allowed environmental support. A real report should define that scoring rule.

The highest percentage occurred during the week with the fewest opportunities and one sampled routine. Compared with May 18, the final week had a lower percentage, more opportunities, the same listed support level, and broader conditions. Compared with the first two weeks, it also omitted the gesture prompt. The pattern could reflect progress, setting differences, sampling differences, or several influences together. The responsible clinician should examine the raw observations and explain the conclusion.

Ask which events were excluded. A missing AAC device, an adult who never created the defined opportunity, a health event, and a child declining a probe have different meanings. When preparing materials is a required staff step, record a failure to prepare them as a treatment-integrity miss. Mark the corresponding child probe as invalid or missed according to the written measurement rule, and report both numbers.

Separate prompts, independence, fluency, and quality

A correct response after a full model differs from a response initiated without an adult prompt. Reports should state the prompt hierarchy, define “independent,” and show whether help is decreasing. A high accuracy score can coexist with prompt dependence.

Some goals also require a measure beyond accuracy. Communication may need to occur soon enough to be useful. A daily-living routine may need safe, comfortable completion at a workable pace. A safety response may require scoring each step in a chain. Quality measures should stay observable and should avoid judging personality, harmless self-regulation, eye contact, or appearance as treatment failures.

Distinguish acquisition, mastery, maintenance, and generalization

These terms answer different questions:

• Acquisition asks whether the skill is developing during teaching.
• Mastery means the child met the plan's stated criterion under the stated conditions.
• Maintenance asks whether the skill continues after time passes or teaching becomes less frequent.
• Generalization asks whether the skill occurs with relevant people, materials, routines, or settings beyond the trained examples.

Mastery in one therapy room with one technician does not establish use at home, school, or in the community. Generalization also should not require the child to perform everywhere regardless of comfort, safety, or preference. Check which conditions were tested, which supports remain useful, and which relevant context the child and family chose next.

Check whether the plan was delivered as designed

Outcome data are hard to interpret when the report omits what actually happened. Look for treatment-integrity or fidelity information: whether planned steps occurred, the dose or number of opportunities delivered, whether staff were trained, and which environmental disruptions affected the plan.

Observer agreement is another quality check. It asks whether two trained observers using the same definition recorded similar events. Agreement can reveal unclear definitions or inconsistent scoring. A high agreement score cannot prove that the selected measure captures the family's priority, that the opportunities are representative, or that the treatment caused the change.

A routine progress graph describes change over time. Establishing that a treatment component caused the change requires an appropriate experimental design and clinical analysis. Wording such as “increased after,” “was associated with,” or “remained variable” fits a design that supports description rather than a causal conclusion.

Reconcile recommended, authorized, scheduled, and delivered services

Service hours need four separate labels. Recommended hours are the clinician's current recommendation. Authorized hours are what the health plan approved for a defined period. Scheduled hours reflect the calendar. Delivered hours are the services that occurred. Show cancellations, staffing gaps, family availability, illness, holidays, and other barriers without assigning blame.

A gap can affect interpretation. If 24 hours were recommended, 20 authorized, 18 scheduled, and 13 delivered in a month, then about 72% of scheduled hours were delivered, delivered hours equaled 65% of authorized hours, and delivered hours equaled about 54% of recommended hours. Each ratio answers a different question. The report should name its numerator and denominator and explain why the amounts differ. A clinical recommendation still needs individualized rationale. An authorization limit alone does not establish the clinically appropriate amount.

Payer requirements vary. As one current, narrow example, the TRICARE Autism Care Demonstration page states that ABA treatment authorizations run for six months, providers request reauthorization every six months, and named outcome measures occur on six- or twelve-month schedules depending on the measure. Those rules apply to that program and should not be generalized to another plan.

The CMS Prior Authorization API FAQ describes administrative exchanges for CMS-0057-F impacted payers. It says a decision response must state whether the payer approves the request and for how long or denies it with a specific reason; when necessary, the response may request more information. The FAQ supplies no universal ABA progress-report template. Families should distinguish a clinical progress finding from a payer request, authorization decision, appeal, or portal status.

Look for the child's experience and the family's priorities

Progress includes more than target frequency. A report should address assent and dissent, communication access, discomfort, adverse effects, meaningful choice, family feasibility, and quality of life when they affect care. Ask how the child experiences the goal and whether the plan preserves useful supports.

A child who uses augmentative and alternative communication (AAC) needs access to that system during assessment, teaching, and independent probes. The ASHA AAC Practice Portal says AAC users should always have access to their communication tools. The record should show when the system was unavailable and should recognize speech, sign, gesture, device use, and other agreed responses accurately.

Changes in pain, sleep, medication, seizures, gastrointestinal health, hearing, vision, mobility, school demands, housing, transportation, language access, staffing, or family routine may alter performance. The BCBA should route medical questions to the appropriate health professional and coordinate within consent, privacy, and each person's scope.

Expect a decision for the next treatment period

The report should end with a decision tied to evidence. Possible decisions include continuing a goal, changing the definition or measurement, adjusting prompts, testing another setting, teaching a prerequisite, reducing burden, coordinating with another professional, fading a mastered target, pausing for health review, or ending a poorly fitted procedure.

For every material change, look for:

• the observation or data that prompted it
• the date the change began
• the child and family's input
• expected benefit and foreseeable burden
• who approved or consented when required
• the measures and review date that will test the change
• a safety, pause, or referral rule when relevant

A payer deadline can set the submission date. Clinical findings should supply the treatment rationale.

Use this red-flag checklist

One unclear item calls for a question. A repeated pattern of missing context deserves a formal review with the responsible clinician or clinical director.

• Goals lack observable definitions or meaningful purposes.
• Percentages appear without counts, opportunities, or measurement rules.
• Graph axes, dates, phase changes, or missing observations are unlabeled.
• Prompted performance is described as independent.
• Mastery is declared from one short sample or one trained condition.
• Flat, worsening, highly variable, or adverse data receive no analysis.
• The report changes a goal or procedure without explaining the reason and family input.
• Staff implementation, cancelled sessions, and unavailable materials are absent.
• Family availability is treated as a character flaw rather than a feasibility variable.
• The child's distress, dissent, pain signals, or communication access are omitted.
• Recommended hours are copied from authorized or scheduled hours without clinical analysis.
• Names, dates, goals, or conclusions appear copied from another record.

Raise immediate safety or health concerns through the child's established clinical or emergency route. Documentation questions can go to the responsible BCBA and clinical director. Coverage questions belong with the plan using the member ID, product, authorization, and current policy. Suspected privacy or record errors should go through the provider's privacy and correction process.

Ask these 15 questions during the review meeting

1. What reporting dates, settings, people, and services does this report include?
2. Why was each current goal chosen, and how does it matter to our child?
3. What exactly counts as a response and a valid opportunity?
4. How many observations support the baseline and the latest result?
5. Which prompts, tools, and environmental supports were present?
6. What do the graph's level, trend, and variability show?
7. Which dates or observations are missing, excluded, or invalid, and why?
8. What changed in treatment, staffing, health, routine, or measurement during this period?
9. How do you know staff delivered the plan as designed?
10. What has been tested for maintenance and generalization?
11. How were our child's assent, dissent, comfort, and communication represented?
12. How did recommended, authorized, scheduled, and delivered hours differ?
13. Which result led to each recommendation for the next period?
14. What would cause the team to continue, change, pause, refer, fade, or end this plan?
15. When will we review the data again, and how can we get an understandable copy?

Families can request an understandable copy

Ask the provider for existing reports, graphs, definitions, and treatment-change notes in a form and format you can use. For a provider subject to the Health Insurance Portability and Accountability Act (HIPAA), an individual generally has a right to access protected health information in a designated record set. HHS explains that a covered entity must provide the requested form and format when readily producible, or an agreed readable alternative when it is not. HIPAA does not require the covered entity to create new explanatory materials or analyses that do not already exist in that record set. Ask separately what plain-language explanation, interpretation, or translation support is available. A personal representative's access follows the scope of authority under applicable law. The HHS right-of-access guidance also explains timing, fees, exceptions, and review rights. Minor-consent, personal-representative, school-record, state, and provider rules can change who may access a particular record.

An understandable report does more than display data. It lets the child and family see what was learned, correct errors, weigh benefits and burdens, and participate in the next decision.