Instructional control of human operant responding during extinction following fixed-ratio conditioning.

College students pressed a button for coins during a fixed-ratio schedule. After 30 minutes the coins stopped, but the button stayed. Before this extinction phase half the group heard: 'No more money will come out.' The other half got no warning.

The researcher counted button presses for 30 more minutes. He wanted to see if a short, honest rule could prevent the usual extinction burst.

Students who were told the payoff was over pressed the button far less. Their response rate dropped quickly and stayed low. The uninstructed group kept pressing hard, showing the classic extinction burst.

One sentence of truth at the right moment cut the burst almost in half.

Zimmerman et al. (1962) extends this lab finding to a real classroom. They also used extinction, but added praise for good behavior. Two emotionally-disturbed boys stopped disrupting once attention was withheld from problem behavior and given for seatwork. The 1970 study shows the power of a warning; the 1962 study shows you can go further and reinforce the replacement skill.

Delini-Stula (1970) ran the same year with similar college students, but used a token economy instead of extinction. Both papers prove operant principles work with adult learners, yet they solve opposite problems: one shuts down unwanted pressing, the other builds accurate quiz answers.

Lydersen et al. (1974) flips the script. Instead of stopping behavior by withholding tokens, they started behavior by handing them out. When reading work earned points, disruption vanished. Together these studies show you can either tell clients reinforcement is over, or make a new behavior earn reinforcement.

Next time you fade reinforcement, tell the client first. A ten-second rule like 'No more stickers today' can spare you a burst of problem behavior. Pair that warning with praise for appropriate responses, just like Zimmerman et al. (1962) did, and you get both fewer bursts and faster skill growth.