Separating response dependency and response-reinforcer contiguity within a recycling conjunctive schedule.

The team built a recycling conjunctive schedule. Every 30 seconds food was ready. The bird also had to peck once inside a tiny window. The window slid early, middle, or late inside the 30-s cycle. They watched how the pigeons timed their pecks.

Only the window spot changed. Food still came every 30 s and still needed one peck. This let them test 'where the work must happen' without changing how often food appeared.

When the peck had to come early, birds pecked fast at the start and then coasted. When the peck had to come late, birds waited and then pecked near the end. Moving the tiny window moved the whole response pattern. The birds followed the rule 'respond where the job is'.

Neuringer (1973) showed that any response-reinforcer link beats no link. M et al. now add that, once the link exists, the exact spot inside the interval still matters. The two papers stack: first show the connection helps, then show location fine-tunes the pattern.

Bacon-Prue et al. (1980) used a DRO to stretch pauses within fixed-interval performance. M et al. stretch or squeeze the burst by sliding the response window. Both prove that tacking a second rule onto an interval schedule reshapes the whole response curve.

Hineline et al. (1969) found contrast in chained schedules: what happens in the first link changes responding in the last link. M et al. show a similar carry-over inside a single interval. The conjunctive rule acts like a mini-chain, and its internal structure steers behavior.

You now know that 'when the client must respond' is its own lever. If you run a FT 30-s plus response requirement for stereotypy reduction, slide the window early to front-load engagement or late to build wait time. One small schedule tweak can shift the whole response wave without touching reinforcer rate.