Preference pulses induced by reinforcement.

Hachiga et al. (2014) watched rats choose between two levers. One lever paid off on a VR-20 schedule. The other lever sometimes stayed on VR-20 and sometimes went into extinction.

The team tracked which lever the rat pressed second-by-second. They wanted to see if a reinforcer creates a short-lived burst of preference for the lever that delivered it.

Right after a pellet dropped, rats kept pressing the same lever for a few seconds. This mini-preference faded faster when the lever entered extinction.

If the lever still paid on VR-20, the brief spike lasted a bit longer. The data show a clear post-reinforcer pulse, even when the lever no longer pays.

Landon et al. (2003) ran a similar rat-lever setup and found bigger pellets make bigger pulses. Together the two studies show both size and extinction matter to the same momentary effect.

Hunter et al. (2019) conceptually replicated the idea with humans and toys. One accidental reinforcer after extinction pulled kids back to a toy, matching the rat pattern.

Cohen et al. (1993) seems to disagree. They saw reinforcers lengthen the current stay but claimed no real preference shift. The clash is timing: L looked at whole-session totals while Yosuke zoomed in on seconds. Both can be true.

Your client may return to a task right after praise, then drift away. That brief surge is the preference pulse. Use it: deliver the reinforcer, then immediately prompt the next response while the pulse is alive. If the task is now harder or unreinforced, move fast—the pulse dies quicker under extinction.