Outcomes & Data Glossary

Begin with a clear question

Data collection is the planned recording of information for a defined purpose. A useful plan names the behavior or experience, setting, opportunity, observer, method, schedule, missing-data rule, privacy controls, and decision the data may inform.

The CASP ABA Practice Guidelines public page places assessment and treatment planning within ABA behavioral health treatment for people diagnosed with autism. Detailed Version 3.0 content is licensed. The public page does not prescribe one data system.

The BACB BCBA Test Content Outline and ABAI basic-principles page cover measurement and evaluation as training or examination content. Qualified clinicians still choose methods for the person and question.

Know what the number measures

Duration is elapsed time from a defined start to a defined end. Rate is a count divided by an observation-time unit. State the unit, because three messages per hour and three messages per ten minutes describe different patterns.

Interval recording samples whether behavior occurred during or at selected parts of an observation. Interval methods can estimate occurrence patterns while introducing bias. Ask whether the graph shows observed intervals, estimated duration, or another derived value.

Counts and percentages should include their denominators. If the team reports “80%,” ask: 80% of which opportunities, across how many days, with which prompts, in which settings, and with which ordinary supports?

Read the display without overclaiming

A graph displays data using defined axes, units, points, and phase information. A trend is the direction and rate of change across observations. Variability, missing points, scale, phase length, and setting changes affect interpretation.

The same data can look different on compressed or expanded axes. Ask to see raw counts, observation dates, phase changes, and the data-collection method. A visual pattern can support a decision without proving what caused it.

Use progress monitoring as a feedback loop

Progress monitoring is repeated review of data to decide whether goals, supports, teaching, measurement, or services should continue or change. It should include the person and family, rather than operating as an automatic rule based on a threshold.

An outcome measure captures a result that matters for the chosen question. It may measure a skill, participation, communication, independence, comfort, burden, health, safety, relationship, or quality-of-life domain. A standardized score and a direct behavior measure can answer different questions.

Define the baseline, desired outcome, review window, meaningful-change rule, risks, and action options before the team sees the result. This reduces the temptation to select the most favorable metric later.

Put quality of life and social validity beside performance

Quality of life is a multidimensional concept involving a person's wellbeing, choice, relationships, participation, health, safety, material conditions, and other valued domains. Research on quality-of-life measurement shows why the construct should not be reduced to one proxy chosen by professionals.

Social validity concerns the importance and acceptability of goals, procedures, and outcomes to affected people. Peer-reviewed discussion of social validity highlights several ways it can be assessed. A satisfaction form is one source, and direct client communication should be included whenever possible.

Ask the person what changed, whether the effort felt worthwhile, which supports should remain, what was uncomfortable, and what they want next. A goal can meet a technical criterion while leaving daily life unchanged or more burdensome.

Combine evidence without blending its authors. The client's report, caregiver observations, clinician measurement, school or workplace information, standardized assessments, health records, and payer data may describe different settings and periods. Label every source, date, context, and limitation. A disagreement can reveal a real context difference, a definition problem, an access barrier, or a change in the person's priorities.

Before changing care, ask whether the measure was valid for the question, whether implementation and ordinary supports were available, and whether enough eligible observations have matured. Keep improvement, stability, deterioration, missing data, and uncertain interpretation as separate states. The next decision may be to continue, teach, adapt the environment, change the measure, seek another professional evaluation, reduce burden, pause, or discharge through a planned process.

Record the decision, its author, and the next review date.

A family data example

Samira is a fictional ten-year-old who wants to take part in weekend baking. Across five eligible steps during baseline, she chooses and completes two with her usual AAC and mobility supports. Across five later steps using a visual recipe, she completes four.

The graph shows 2/5 and 4/5 under different phases. It should also show prompts, available supports, step difficulty, missing opportunities, and Samira's view. The small before-and-after comparison does not establish that the recipe caused the difference. It can still guide a shared decision about what to try next.

Find ABA care near you. Ask a Finni provider how goals, data, quality of life, client feedback, and progress decisions are handled.