A Pilot Investigation of Individual and Dyad Instructional Arrangements

The team compared two ways to run discrete-trial training: one child with one adult, or two children with one adult.

Four autistic children worked on naming items and following directions. The adults flipped the two formats every day so each child served as his own control.

Sessions were videotaped and scored for how many targets each child learned and kept after two weeks.

Three of the four kids learned more words and kept them longer when they were taught alone.

The same three kids also spent less total time in teaching when they were paired with a peer.

One child did about the same in both set-ups, so the results were mixed overall.

Hassin-Herman et al. (1992) showed that kids with autism can move from 1:1 to small groups without losing skills. Their groups had three to five learners, while Croner used only dyads. The larger groups still kept academic gains, hinting that adding more peers may help, not hurt.

Knopp et al. (2023) used the same alternating-treatments design to compare telehealth and in-person DTT. Both formats worked equally well for expressive labels. Together the studies say the delivery mode—online, alone, or with one peer—matters less than picking the right match for each learner.

Haq et al. (2019) swapped the tabletop for play-based embedded DTT and still saw good skill gains. Pairing their tweak with Croner’s seating plan could let you keep high acquisition while cutting problem behavior and peer distraction at once.

If a child is flying through targets solo, stay individual. If time is tight and the child tolerates peers, try a dyad first, then probe alone after a week to check maintenance. Track trials to criterion each day so you can flip back quickly if learning slows.