The effect of timeout on performance on a variable-interval schedule of electric-shock presentation.

Green et al. (1975) worked with lab subjects who pressed a key for food on a VI schedule.

Every press also gave a quick electric shock.

After each shock the chamber lights went off for 30 seconds.

The team then removed or shortened this blackout to see if pressing would stop.

When the blackout disappeared, key pressing almost stopped.

Cutting the blackout to 5 seconds also cut responding.

The blackout, not the shock, was keeping the behavior alive.

Timeout was acting as a reinforcer, not a punisher.

Dunham et al. (1969) saw the opposite with pigeons: timeout after food suppressed fast pecks.

The difference is the shock.

With food, timeout removes good stuff and acts like a punisher.

With shock, timeout removes bad stuff and acts like a reward.

Last et al. (1984) later showed shock alone can maintain behavior when it is given right after long pauses.

Together the three papers show the same timeout can punish or reinforce depending on what it removes.

If you use timeout after aversive events, check what it removes.

A kid who hits to escape noise may hit more if your timeout also removes noise.

Try giving escape another way, like a break card, and reserve timeout for times it truly removes something good.