Undermatching and contrast within components of multiple schedules.

McLean et al. (1981) watched pigeons peck keys under two multiple schedules. Each schedule had rich and lean components that swapped every few minutes.

The team counted pecks in ten-second bins. They wanted to see how fast matching faded after each swap.

Right after a component switch, birds' response ratios almost matched the new reinforcer ratio. Sensitivity then slid downhill for the rest of the component.

Local contrast popped up early too, but its direction flipped depending on whether the upcoming component stayed the same or changed.

Carr et al. (1985) later showed the same downhill slide inside longer components, proving the fade is real, not a quirk of short timing.

Szatmari et al. (1994) swapped regular and irregular alternation and found no change in contrast or matching. Together the papers say: what matters is time since transition, not whether the bird can predict it.

Scull et al. (1973) saw positive contrast only when both components used the same response. The 1981 data echo that: contrast direction hinges on context details, not just reinforcer rates.

If you run multiple schedules in therapy, probe early and late within each component. Early data give the cleanest picture of true reinforcer control; late data may hide undermatching. When you write session notes, flag the first 30 seconds after any schedule switch as the window most likely to show sensitivity. If contrast looks weak, check whether the response form stays the same across components before you tweak reinforcer rates.