Resurgence of previously taught academic responses

Williams et al. (2020) taught college students to solve quadratic equations step-by-step. Later they reinforced a new way to solve the same problems. Finally they stopped all rewards. The team watched to see if the first method popped back up.

Sixteen students took part. The key twist: some extinction problems looked like the original training set, others did not. The researchers counted how often the old math chain returned.

The first-taught equation chain came back for ten of the sixteen students. Resurgence was strongest when the extinction problems resembled the original target. Context mattered.

Old academic skills do not disappear; they wait. When the new strategy quits paying off, learners reach back for the earlier one—especially if the worksheet looks familiar.

Garner et al. (2018) saw the same rebound in kindergarten readers. Phonetic sounds returned when reinforcement stopped. Together the two studies show resurgence is not just a lab curiosity—it hits both college math and early literacy.

Craig et al. (2017) and Shahan et al. (2020) found bigger resurgence when the alternative reinforcer was large before extinction. Williams adds a new lever: problem format. Similar worksheets wake up the old response more than novel ones.

Greer et al. (2024) extended the rule to destructive behavior. Big early cuts in reinforcement produced the largest relapse. Williams mirrors this in academics: the more the extinction task resembles training, the bigger the bounce-back.

When you fade an old skill, watch the worksheet, flashcard, or app screen. If it looks like the original training set, the learner may slide back into the previous strategy. Swap to fresh materials during extinction or thinning. This small context change can cut resurgence before it starts.