The PEAK assessment in ABA: what it measures and how teams use it

The PEAK assessment is the assessment half of the PEAK Relational Training System (Promoting the Emergence of Advanced Knowledge), an ABA curriculum for language and cognition by Mark Dixon. Each PEAK module has an assessment that shows which skills a learner already has, plus programs to teach the missing ones: assess, teach, then assess again.

PEAK is used by behavior analysts with children and adults with autism and other developmental disabilities, often after early VB-MAPP levels are mastered.

The PEAK assessment is based on three lines of behavior science: Skinner’s analysis of verbal behavior for the first two modules, and stimulus equivalence and relational frame theory (RFT) for the last two. Relational frame theory studies how people derive new relations, such as “bigger than” or “same as”, without being taught each one.

A simple example: a child learns that the spoken word “dog” goes with a picture of a dog, and that the picture goes with the printed word. If the child then matches the spoken word to the printed word without ever being taught that pair, that is derived relational responding. PEAK tries to teach that kind of flexible learning on purpose.

The four PEAK modules are Direct Training (PEAK-DT), Generalization (PEAK-G), Equivalence (PEAK-E) and Transformation (PEAK-T). Each PEAK module lists 184 target skills, so the full system covers 736 programs, usually worked through in order, although a learner can be in more than one module at once.

The PEAK Comprehensive Assessment (PCA) is one standardized direct assessment of language and cognition that pulls items from all four PEAK modules, with guidance on which programs to start. Before the PCA, a full PEAK workup meant running each module’s assessment separately, which takes far more time with the learner.

The PCA does not cover every one of the 736 programs. It samples them. Research on the PCA reports a strong link between PCA scores and the module pre-assessments, but it is still a sample. When a learner is close to a gap, many teams check the specific programs directly before teaching.

PEAK scoring works in layers within each module: an indirect assessment where a parent or teacher answers yes, no or don’t know for each skill, a direct pre-assessment often run with a flip book, and short trial blocks for skills that remain unclear. The PEAK module score is essentially a count of skills shown.

The result is a picture of which of the 184 skills are in place and which are not. A skill that is missing becomes a candidate program, and because the score is a count, you can compare it over time. Daily progress on each program can be logged on free ABA data sheets.

The PCA gives a combined picture across the four modules. Use the official manual for the exact scoring rules, scoring sheets and how scores translate into program choices.

The PEAK assessment has no fixed age range: PEAK was built for children and adults with autism and other developmental disabilities, and what matters is the learner’s skills. The Direct Training module starts with very early skills, while the Equivalence and Transformation modules reach into reasoning that typically developing school-age children show.

Teams pick PEAK when they want early skills and later reasoning skills in one system, or when a learner has mastered early VB-MAPP levels and needs more flexible, untaught responding.

PEAK and the VB-MAPP overlap most in early skills: the PEAK Direct Training module covers requesting, imitation, matching to sample and listener skills that the VB-MAPP also measures. The difference is reach: the VB-MAPP stops at about 48 months of language development, while PEAK continues into generalization and derived relational responding.

SuperChild’s trainings are built from VB-MAPP milestones and ABLLS-R tasks, so many early Direct Training skills already have practice sets there; the professional picks any that fit a PEAK target. For table work, matching activities and feature, function and class activities cover several early PEAK-DT and PEAK-G targets. For a side-by-side of all five tools, see the ABA assessment comparison.