The effects of the question "What is this?" on tact-training outcomes of children with autism.

The team taught new object names to four children with autism. They ran discrete-trial tact sessions two ways: some trials started with the adult asking “What is this?” and some did not.

An alternating-treatments design flipped the two styles across days. The goal was to see if the extra question sped up learning or kept the new words stronger later.

Two kids learned the names faster when the adult asked “What is this?” first. The other two learned faster without the question. A month later, all four kept the words equally well no matter how they were taught.

In short, the question helped some and hurt others, but it did not change long-term memory.

Lloyd et al. (2021) saw the same child-to-child swing when they tested four error-correction styles. One kid did best with model-lead, another with echoic, and so on. Both studies tell the same story: small procedural tweaks matter, but the winner is different for every learner.

Dell’Aringa et al. (2021) compared transfer trials to plain DTT and found no overall speed gain. Marchese et al. (2012) adds a second null at the group level: the question did not win across the board either. Together, the papers warn us not to pick one “best” format before testing it with the individual.

McCormack et al. (2017) looked at a different tweak—pairing each correct tact with a unique toy sound or snack. They also saw a split: two of three kids learned faster with the special reinforcer, one did not. The pattern echoes V et al.: when you change one variable, some kids sprint and some slow down.

You already probe prompts and reinforcers kid-by-kid. Do the same with the question “What is this?”. Run a quick five-trial mini probe both ways, track correct responses per minute, and pick the faster slope for that child. No extra materials needed—just your timer and data sheet.