Effects of serial and concurrent training on receptive identification tasks: A Systematic replication

The team taught preschoolers with autism or developmental delay to pick letters or letter sounds. They compared two ways to run trials: serial (one letter at a time) or concurrent (three letters shown together).

Each child got both styles in an alternating pattern until they hit mastery. The researchers tracked how many trials and days each kid needed.

Concurrent training shaved off a few sessions, but the gap was small. Both groups reached mastery and could point to new letters in untaught sets.

Kids learned at nearly the same speed whether letters popped up alone or in groups of three.

Vladescu et al. (2021) also used alternating treatments and found smaller sets (3-6 items) beat large sets (12) for adolescents learning tacts. Wunderlich’s work echoes that point: keeping the set modest helps, but going from one to three items gives only a tiny boost.

Cordeiro et al. (2022) asked a different question—how to decide mastery—yet used the same quick-switch DTT design. Their finding that target-level mastery halves teaching time pairs nicely here: you can run concurrent trials AND judge mastery per single stimulus to save even more time.

Berkovits et al. (2014) taught letter-sound skills through play, not tabletop trials, and still saw gains. This reminds us that the serial-vs-concurrent debate matters most when you are already at the table; naturalistic teaching can work too.

If you run discrete trials for receptive labels, you can stop worrying about perfect trial order. Present one item or three—both work. Use concurrent when you want a slight speed edge or to keep the learner’s attention with more visual variety. Pair this with target-level mastery from Cordeiro’s paper and you may cut total sessions without hurting generalization.