A Component Analysis of Skill Acquisition Consequences with Listener Responses

Zhi et al. (2024) split listener-response DTT into three parts. They tested praise-only, correction-only, and the full learn unit together.

An alternating-treatments design let each learner sample every style in one session. The team tracked correct listener responses and how long the skill stuck.

Correction-only matched the full learn unit. Both beat praise-only for teaching and keeping listener responses.

Adding praise to correct answers gave no extra boost. Quick, clear error fixes were enough.

Osnes et al. (1986) saw the same pattern in positive-practice overcorrection. Reinforcing correct practice sped learning, but the later study shows praise is not required.

Kangas et al. (2011) warns that even half-wrong corrections ruin DTT. Zhi et al. adds that once the correction is accurate, praise is optional.

McGee et al. (1983) showed shorter 30-second positive practice works as well as long bouts. Together these papers say lean, accurate corrections are both effective and efficient.

You can save your breath. Deliver a quick, firm correction after an error and move to the next trial. Skip the extra praise if time or attention is low; the learner will still master listener responses. Try it next session: run five correction-only trials, five full-learn-unit trials, and plot the data. You may find equal learning with less effort.