Temporal proximity and reinforcement sensitivity in multiple schedules.

Carr et al. (1985) asked a simple question: do shorter schedule parts make pigeons more sensitive to reinforcement? They ran two multiple VI schedules side by side. One schedule switched every 15 seconds. The other switched every 60 seconds.

The team measured how well response ratios tracked reinforcer ratios at three moments inside the 60-second parts: right after the switch, in the middle, and near the end.

Fifteen-second and sixty-second parts gave the same sensitivity scores. Short parts did not boost sensitivity.

Within the 60-second parts, sensitivity started high then dropped the longer the bird stayed in that part. Timing of the probe mattered more than part length.

Reiss et al. (1982) looked like they disagreed. They saw higher response rates in short parts. But they also paired short parts with richer reinforcement. When they reversed the pairings, the effect vanished. Their data and G et al. now agree: part length itself does not change sensitivity; correlated reinforcer rate does.

McLean et al. (1981) had already shown that sensitivity peaks early in a part then falls. G et al. replicate and sharpen that picture. They show the decline is steady, not just a quick post-switch burst.

Rutter et al. (1987) later confirmed the main message. Even when part changes depend on number of responses instead of time, reinforcer rate still drives the response pattern. Part duration rules add little extra force.

When you run multiple schedules in a functional analysis or treatment evaluation, probe early in each part if you want the clearest sensitivity measure. Do not assume that shortening the part will automatically strengthen control by reinforcement. Check whether richer reinforcement is accidentally paired with shorter parts before you credit the duration change. In practice, keep part length consistent and measure early to avoid underestimating sensitivity.