Time and rate measures in choice transitions.

Researchers watched pigeons peck two keys under concurrent and chained schedules. They timed how long birds waited before the first peck and counted pecks per minute.

Three small lab tests mapped these wait-time and rate measures as schedules changed. The goal was to see if the two measures move together or take turns.

Wait times and response rates formed clean straight lines or gentle curves. When schedules switched, changes in the first link lagged behind changes in the final link.

The birds did not flip choices all at once. Time data showed the shift earlier than rate data did.

Szempruch et al. (1993) saw pigeons stretch or shrink wait times when food timing changed, but the birds lost track after many switches. Cerutti et al. (2004) add the missing piece: even after tracking fades, the link between wait time and rate stays linear.

Nevin (1969) split fixed-interval responding into two states—pause then burst. The new study shows the same two-step inside choice transitions, but now with math curves, not just labels.

Miranda-Dukoski et al. (2014) used ever-shifting food odds and found choice tracking fades unless brief signals are given. T et al. used steady schedules and still found lag; signals are not needed when the rules stay put.

Sanders et al. (1971) showed choice ratios match reinforced inter-response times. The 2004 data keep the matching rule, yet reveal that wait-time clocks the change before the rate counter does.

If you only graph responses per minute, you can miss when the learner first notices the schedule shift. Add a simple stopwatch measure—time until the first response—and you may catch early discrimination that rate hides. Try it during concurrent-schedule probes or when you thin reinforcement. The lag signal can tell you to wait a bit longer before you raise the demand again.