An analysis of timeout and response cost in a programmed environment.

The team worked with six teens who had intellectual disability. All lived in a treatment center.

Each teen earned tokens for good behavior. The staff could also take tokens away or place the teen in a quiet room.

The study compared small and large consequences. Five tokens lost versus thirty. Five-minute timeout versus thirty. Staff rotated the conditions to see which cut antisocial acts the most.

Thirty-minute timeout and thirty-token loss beat the five-unit versions almost every time.

Problem behavior kept dropping under the big consequences. The small ones worked at first, then effects faded.

Leander et al. (1972) ran a near-copy study the same year. They also found thirty-minute timeout worked, but saw no extra gain over fifteen minutes. Together the papers show: at least fifteen minutes is needed, but thirty may give a safety margin.

Capio et al. (2013) later asked, "What if kids can shorten timeout by behaving?" Preschoolers complied more when good behavior trimmed the timer. Their result does not cancel the thirty-minute rule; it simply adds a classroom trick you can layer on top.

Van der Molen et al. (2010) tested response-cost with adults decades later. Money loss still suppressed button pressing, proving the principle survives age, setting, and type of token.

When you need fast suppression of serious behavior, go big. A half-hour timeout or a steep token fine gives stronger, longer-lasting drops than a slap-on-the-wrist five. Keep the smaller options for mild misbehavior or when safety rules cap duration.