Assessing the repeatability of resurgence in humans: Implications for the use of within‐subject designs

Kestner et al. (2018) asked a simple question. If you test resurgence twice in the same person, does it come back just as strong?

Adults pressed buttons for points on a computer. First, pressing the left button paid off. Then only the right button paid. Finally neither button paid. This last part is extinction. The left-button presses that return are resurgence.

Two weeks later the same adults repeated the whole sequence. The team compared the size of resurgence in round one versus round two.

Resurgence shrank the second time. The first extinction gave a big burst of old button presses. The second gave barely a blip.

The drop was large enough to wreck within-subject designs. You can’t fairly compare two treatments if the mere order changes the outcome.

Nist et al. (2021) saw the opposite. They also ran two extinction tests, but resurgence stayed strong across both. The difference? They thinned the alternative reinforcement within each session. Kestner removed it all at once. Method shapes outcome.

Ritchey et al. (2021) also repeated extinction with humans and still got resurgence. Yet they lengthened initial training before each test. More training history can overpower the order effect, keeping relapse alive.

Greer et al. (2024) worked with children who had severe behavior. They found minimal resurgence when thinning was gradual. Again, the way you withdraw reinforcement matters more than how many times you test.

If you plan to compare two resurgence-reduction tactics in the same client, don’t alternate them back-to-back. The second round will look better even if it isn’t. Instead, use separate groups or give enough time and re-training between tests. Always taper alternative reinforcement slowly; abrupt stops invite bigger, less reliable relapse.