ABA Fundamentals

Units of analysis in acquisition‐performance criteria for “mastery”: A systematic replication

Wong et al. (2022) · Journal of Applied Behavior Analysis 2022
★ The Verdict

Check mastery per single item, not the whole set, to cut sight-word training time with no loss in retention.

✓ Read this if BCBAs running discrete-trial or table-time lessons with any learner working on sight words, tacts, or listener discriminations.
✗ Skip if Instructors already using individual-item mastery rules or teaching solely with larger set-size protocols shown effective in other studies.

01Research in Context

01

What this study did

Wong et al. (2022) asked a simple question: should we call a skill "mastered" when every single item hits criterion, or when the whole set averages 100%?

They taught sight words in discrete trials. In one condition the learner had to get each word right nine times out of ten before moving on. In the other condition the whole set of words had to average 100% correct.

Three participants alternated between the two mastery rules while the team counted trials and checked next-week retention.

02

What they found

Individual-word mastery won. All three participants reached the final performance goal in fewer trials when each word had to pass on its own.

Maintenance scores one week later were the same for both rules, so the faster route did not cost long-term recall.

03

How this fits with other research

Chang et al. (2024) ran the same comparison with four children and got the same result: checking each tact individually cut trials to criterion. The replication two years later strengthens the claim.

Cordeiro et al. (2022) extended the idea to tacts, intraverbals, and listener responses in kids with autism. Again, target-level mastery beat set-level mastery, showing the rule works across skill types.

Kodak et al. (2020) seems to disagree. They found bigger stimulus sets (6–12 items) produced faster tact learning than smaller sets. The difference is the mastery unit: Kodak kept the unit at the set level, while Wong moved it to the individual item. Smaller sets plus individual mastery is the sweet spot.

04

Why it matters

If you wait for an entire set to score 100% before advancing, you may waste trials on items already mastered. Flip the criterion to the single item and you let mastered words drop out early, shrinking total teaching time. The evidence now spans sight words, tacts, intraverbals, and listener skills across typical learners and children with autism. Next time you write a mastery check, require 90% on each target, not the whole set, and watch the session count fall without hurting maintenance.

Free CEUs

Want CEUs on This Topic?

The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.

Join Free →
→ Action — try this Monday

Rewrite your data sheet so each word must hit 9/10 correct before it is removed from rotation; stop waiting for the entire set to average perfect.

02At a glance

Intervention
discrete trial training
Design
alternating treatments
Sample size
3
Population
not specified
Finding
positive

03Original abstract

This study compared 2 units of analysis for assessing acquisition mastery during sight word instruction for 3 participants. The unit of analysis refers to the specific performances that criteria are applied to, either sets of stimuli or individual operants. In the Set Analysis condition, we applied the acquisition-performance criterion to the aggregated accuracy of a set of 4 target operants. In the Operant Analysis (OA) condition, we assessed the criterion for individual operants and replaced targets as they met the acquisition criterion. All participants acquired novel textual responses to sight words faster under the OA condition and response maintenance was similar between conditions. This study extended previous research by showing enhanced response maintenance in the OA condition by increasing the performance criterion from 1 observation of 100% accuracy to 2. This study also suggests a unique contribution of OA to quickening learning.

Journal of Applied Behavior Analysis, 2022 · doi:10.1002/jaba.915