Sorting test as a measurement of expansion of equivalence classes
A fast picture-sort after simple training shows if equivalence classes have grown.
01Research in Context
What this study did
The team asked 30 college students to sort pictures into piles. First the students learned two simple pairs: A goes with B, B goes with C. Then new pictures showed up. The researchers watched if the students put the new items in the same piles.
After the sort, the students did a regular match-to-sample test on the computer. The whole session took less than one hour.
What they found
Eight out of ten students sorted the new pictures into the correct piles. That quick sort predicted who would pass the computer test.
Only two students failed both the sort and the computer test. The sorting task caught class expansion without extra training.
How this fits with other research
Reynolds et al. (2022) also got big gains in one hour. They taught absolute pitch with a theremin. Both labs show that brief, clear drills work fast with adults.
Bacotti et al. (2021) looked at feedback timing. Their adults changed what they liked as they got better. Arntzen’s adults also shifted: first they learned simple pairs, then they showed full classes. Both studies remind us to check again after the skill grows.
Ribes-Iñesta (1999) tells us Thorndike cared about transfer. Sorting is modern transfer: students take a new picture and treat it like the old ones. The old idea lives in the new task.
Why it matters
You can use a two-minute sort instead of a long match-to-sample test. If the learner sorts new items into the right piles, the class has likely expanded. Save time, run the sort, then teach the next set.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →After teaching two trained pairs, hand the learner five new pictures and say, "Put together the ones that go together." Count correct piles as your probe.
02At a glance
03Original abstract
The primary purpose was to study how the expansion of equivalence classes is documented by sorting tests. In two experiments with 40 adult participants, there were three phases of training and testing of emergent relations. In the first phase, the participants were trained on 12 conditional discriminations arranged as a linear series training structure (A➔B➔C➔D➔E) followed by a sorting test. The second phase included simple discrimination training of C stimuli. The training comprised different numbers of key presses, and these numbers were used as F stimuli in the expansion test of the existing classes. The final phase contained sorting and matching-to-sample (MTS) tests. The two experiments differed in the number of key presses in the simple discrimination training and stimuli used as F stimuli in Phase 2 and the order of sorting and MTS tests in Phase 3. The main findings of the two experiments were that 100% of the participants sorted the stimuli correctly in the first phase, 83% (25 of 30) of the participants showed expansion and sorted the stimuli in the second phase, and finally, 90% (36 of 40) of the participants responded correctly on the MTS test in the third phase.
Journal of the Experimental Analysis of Behavior, 2026 · doi:10.1002/jeab.70085