Sequential reacquisition as a function of timeout from avoidance.

Researchers tested how long a timeout should last during an avoidance task. They used lab animals that could press a lever to avoid mild shocks. After each session, they gave a signaled timeout break.

They tried timeouts ranging from a few seconds to several minutes. Then they measured how fast the animals responded and how often shocks still happened.

Timeouts of 15 to 60 seconds gave the best results. Animals responded quickly and got fewer shocks. Very short timeouts (under 15 s) or very long ones (over 60 s) made performance worse.

Surprisingly, the error rate stayed the same no matter the timeout length. Only speed and shock count changed.

Davis et al. (1974) showed timeout can accidentally reward mistakes. Their animals kept making the same errors when timeout followed them. The current study flips that idea: here timeout was planned, not a mistake, and it helped when timed right.

Rutter et al. (1987) later proved timeout can act like a reinforcer on a VI schedule. They built on the 1976 parametric work by showing the effect holds across different schedule types.

Solnick et al. (1977) replicated the core idea with children. They found timeout only works if the regular room is more fun than the break room. Together, these four papers show timeout is never neutral; it always functions as a consequence, good or bad.

When you use timeout, think duration first. Aim for about 15–60 seconds for most clients. Shorter gives no relief; longer wastes time and can weaken motivation. Always check whether timeout is actually punishing or accidentally reinforcing the behavior you target.