The power of one reinforcer: The effect of a single reinforcer in the context of shaping

College students played a computer game with two objects. One object had a long history of paying off with points. The other object had just been put on extinction.

After the player stopped touching the dead object, the experimenter gave one lone reinforcer. Then they watched where the player spent time next.

That single point was enough to make players hang around the dead object as long as they did with the rich one.

One accidental payoff erased the extinction effect and pulled behavior back to the wrong target.

Hachiga et al. (2014) first saw this in rats. After one pellet, the rats returned to the lever that was now in extinction. The new study shows the same bounce-back happens with humans.

Cohen et al. (1993) said one reinforcer only gives a brief local spike. Their birds stayed on the rich key in the long run. Hunter’s humans, however, kept preferring the once-dead object for the rest of the session. The difference is timing: L looked at steady concurrent schedules, while Hunter used shaping and a long extinction pause before the lone payoff.

Jimenez-Gomez et al. (2025) add that reinforcers work like traffic signals. They tell the learner what is likely to pay next. The desperation-driven click is powerful because the learner reads it as "this spot is hot again."

If you are shaping a new response, one mistimed treat can lock in the wrong form. Wait for the exact topography you want before you click again. Check your data after each session; a sudden jump in old errors may be the footprint of an accidental reinforcer.