Comparing Skill Acquisition Under Different Stimulus Set Sizes With Adolescents With Autism Spectrum Disorder: A Replication

Two teens with autism learned to name pictures in discrete-trial sessions.

The teacher compared three set sizes: 3 cards, 6 cards, or 12 cards per lesson.

An alternating-treatments design rotated the sizes across days while keeping everything else the same.

Both kids mastered the 3-card and 6-card sets faster than the 12-card set.

Sessions with 12 cards needed almost twice as many trials to reach mastery.

Error rates stayed low in the small sets but climbed when the pile grew to 12.

Kim et al. (2023) also ran alternating treatments and found that 12-trial mini-sessions beat 20-trial marathons for sight words. Together the papers say: keep the load light—fewer items and shorter runs speed things up.

Kodak et al. (2022) used the same design to test mastery criteria, not set size, and likewise showed that leaner conditions win. The matching method gives us confidence the result is real, not a fluke.

Rapp et al. (2016) review task interspersal—mixing easy and hard tasks. Their call for tighter parametric work is answered here: Vladescu et al. (2021) pinpoint the sweet spot at 3-6 exemplars before interspersal gains fade.

Next time you build a tact program, cap the set at six pictures. You will cut teaching time almost in half and see fewer errors during acquisition. If you have 12 targets on the plan, split them into two smaller sets and rotate; the learner masters faster and you keep momentum high.