On the predictive validity of behavioral momentum theory for mitigating resurgence of problem behavior

Fisher’s team ran four small lab studies with adults who had developmental disabilities.

They taught each person a new communication response. Then they thinned the rewards.

Before thinning, some responses got lots of reinforcement. Others got only a little.

Later the team stopped all rewards to see if the old problem behavior would return.

Responses that earned many rewards before thinning stayed strong during the extinction test.

Old problem behavior returned less often for those responses.

In plain words, dense reinforcement first acted like a shield against resurgence.

Craig et al. (2019) saw the opposite: repeated extinction made behavior weaker, not stronger. The difference is Craig kept exposing the same response to extinction, while Fisher built a rich history first.

McIntyre et al. (2002) already showed rich histories make behavior tougher. Fisher moves that idea from basic pigeon work to clinical resurgence.

Silva et al. (2025) add another layer. They found fading the therapy room back in cuts renewal. Fisher’s momentum trick and Silva’s context fade can be stacked: thicken reinforcement, then fade the room slowly.

Before you thin reinforcement, juice up the replacement response. Give 5–6 quick rewards for every one you plan to keep. Then thin. This simple front-loading can spare you a later burst of problem behavior when extinction hits.