The effects of intertrial interval and instructional format on skill acquisition and maintenance for children with autism spectrum disorders

Cariveau and team compared four ways to run discrete-trial teaching.

They varied the gap between trials (2 s vs 5 s) and the trial order (same question block after block, or mixed questions).

Kids with autism got short teaching bursts on naming items and answering questions.

The 2-second gap with mixed questions won.

Kids reached mastery faster and kept the skills later.

Same-length gaps in blocked drills helped less.

Sanford et al. (1980) showed the same thing 36 years earlier: shorter gaps speed learning.

Cariveau adds the twist that mixing trial types matters too, so the old rule now has a partner rule.

Cox et al. (2015) looks like a clash—short gaps made rats and pigeons pick tiny, quick rewards.

The animal study used choice tasks, not skill teaching, so the “impulsive” effect doesn’t apply here.

You can shave minutes off each program by cutting the pause to two seconds and shuffling your questions.

No extra cost, no new materials—just faster mastery and better maintenance.

Try it in your next tact or intraverbal set and watch the data climb.

The intertrial interval (ITI) is the brief pause between the end of one discrete trial and the start of the next. In discrete trial training (DTT), a trial runs from the instruction (the SD) through the learner's response and the consequence; the ITI is the short gap before the next instruction begins.

The purpose of the ITI in DTT is to mark a clean separation between trials. It signals that one learning opportunity has ended and resets the learner for the next discriminative stimulus, so responses stay tied to the intended cue rather than blurring together. ITI length is a small procedural detail that turns out to affect how quickly and efficiently a learner acquires skills, which is exactly what this study set out to test.

The authors replicated and extended an earlier study by comparing three ITI durations, short (2 seconds), progressive (2 to 20 seconds), and long (20 seconds), for two children with autism spectrum disorders. They crossed ITI length with instructional format, comparing massed-trial presentation (the same target repeated) against varied-trial presentation (targets interspersed), while teaching tacts and intraverbals.

Short ITIs in a varied-trial format produced the most efficient skill acquisition. More broadly, most ITI durations led to more efficient learning in a varied format than in a massed format, so trial format mattered a great deal. Which specific ITI and format produced the least stereotypy and the best maintenance and generalization differed across the two participants.

The clinical implication is twofold. First, a short intertrial interval combined with varied, interspersed trials is a strong default for efficient acquisition. Second, because the lowest problem behavior and best maintenance were participant-specific, practitioners should still individualize ITI length and format using each learner's own data rather than assuming one setting fits everyone.