Some effects of response cost upon human operant behavior.

WEINELong (1962) asked adults to press a button for points. The points bought candy later. The team added a twist: each press also cost one point. They tested this response cost on two schedules—VI and FI.

Sessions ran in a small lab room. A counter showed the score. The order of cost vs no-cost conditions flipped for different people so the team could see sequence effects.

Response cost cut VI rates almost in half. On FI, the same cost created clear scallops—long pauses then fast bursts. Without cost, FI responding stayed steady; no scallops appeared.

When cost was removed, rates bounced back. Switching the order of conditions changed the size of the drop but not the pattern.

Lea et al. (1977) moved response cost from the lab to a prison token economy. They found that without close supervision staff took away too many tokens too often. The lab pattern—cost suppresses behavior—held, but real-world use needed oversight.

Iwata et al. (1990) kept the VI schedule and removed cost. They showed that whether participants could earn points outside the session (an open economy) also changed rates. WEINELong (1962) and Iwata et al. (1990) together tell us both cost and outside access shape VI performance.

Kuroda et al. (2018) later showed that the correlation between responses and reinforcers matters even when delay stays the same. WEINELong (1962) adds that the correlation between responses and losses matters just as much.

If you use response cost in a token store, classroom, or home program, expect an immediate drop in behavior and watch for new patterns like long pauses. Check that staff do not over-use cost when you step away—Lea et al. (1977) proved supervision is essential. Finally, test behavior in the real economy; if the learner can earn back-up reinforcers elsewhere, the suppressive effect may shrink.