Formation of new stimulus equivalence classes by exclusion

Plazas et al. (2018) asked if people can build new stimulus classes with only exclusion trials. No praise. No tokens. Just 'pick the one that does not belong.'

They ran three small lab experiments with college students. Each student saw groups of abstract shapes on a screen. In every trial, three pictures came from one class and one picture came from another class. The task was simple: click the odd one out.

Most students ended up treating the never-reinforced 'correct' pictures as if they were old friends. When later tested, they matched pictures that had never been trained together. New equivalence classes had formed from exclusion alone.

The classes held even though the students never got a 'yes' or 'no' after any response. The act of rejecting the odd stimulus was enough to glue the remaining items together.

Fields et al. (2021) extend the idea by adding a timing twist. They also built classes with few trials, but separated the comparison display from the response window. Their tweak doubled success rates, showing that exclusion works even better when you give the learner a brief memory pause.

Ayres‐Pereira et al. (2025) push the boundary in a different direction. They formed classes with almost identical pictures, something earlier work said was nearly impossible. Their trick was showing the near-twins side-by-side during comparisons. Together, the three studies chart a continuum: exclusion creates classes, timing strengthens them, and simultaneous viewing rescues them when stimuli are highly confusable.

Older studies like LeFrancois et al. (1993) used rich compound cues or meaningful pictures to help classes form. Plazas et al. (2018) prove you can dump all that extra support and still succeed—if you let the learner rule out the wrong item each time.

Exclusion trials are already in your toolkit during conditional-discrimination teaching. Now you know they can do double duty: build the class while you test it. Next time you run an identity-matching program or an auditory discrimination task, slip in occasional exclusion probes. You might watch new relations emerge without extra direct teaching, saving valuable session minutes and reducing learner fatigue.