Effect of signaled reinforcement on response variability

Reed tested pigeons in a lab box. The birds pecked keys under two rules.

One rule needed eight different response patterns before food (lag-8). The other needed slow, spaced pecks (DRLF).

Half the time a brief light came on right before food. Half the time food arrived with no signal.

When the lag-8 rule was in force, the signal made the pigeons vary their pecking more.

When the DRLF rule was in force, the signal also made the birds switch up their timing more.

But when no variability was required, the same signal cut variability instead.

Dugdale et al. (2000) first showed you can reinforce variability itself. Reed (2023) adds a twist: a quick flash before food makes the effect stronger.

WEISSMAN (1963) found that signals can weaken schedule control if the time window is too tight. Reed used longer windows, so the signal helped rather than hurt.

Together these papers show timing matters. Short windows scramble control. Long windows plus a signal sharpen it.

If you are shaping flexible responding in a client, try adding a brief cue right before the reinforcer. The cue tells the learner "that varied response was the right one." This works for new play actions, varied vocal phrases, or different ways to solve a task.