Revisiting the use of two choices in the matching‐to‐sample procedure

Boldrin et al. (2022) asked a simple question: do we really need six choices in matching-to-sample, or will two work just as well?

They ran three small lab experiments with college students. Everyone got the usual A→B and B→C training, then tests for symmetry, transitivity, and full equivalence.

The trick: one group picked from two comparisons, the other from six. The wrong pictures changed every trial so learners could not just reject the same foil.

Both groups formed equivalence classes equally well. Two choices were just as good as six.

Rotating the incorrect picture removed the old "reject-control" worry that had pushed clinicians toward bigger arrays.

Fields et al. (2021) showed that splitting the response window from the comparison display can double success. Boldrin adds another easy tweak: cut the array to two, but keep the foil fresh.

Ayres-Pereira et al. (2025) found that near-identical stimuli must be shown side-by-side or classes fail. Boldrin’s rule still holds—just make sure the single foil changes each trial.

Tenneij et al. (2009) cut errors by showing comparisons five seconds before the sample. Boldrin achieves the same goal—fewer errors—by simply shrinking the set size.

Together the papers say: equivalence training is robust; small layout or timing fixes, not bigger arrays, are what help learners win.

You can now run equivalence lessons with only two pictures on the screen. Fewer choices mean faster trials, less scrolling, and easier programming on tablets. Swap the wrong picture every trial so the learner can’t just rule it out by memory. This keeps the task tough enough to teach the relation without overwhelming the student or the instructor.