The effects of simultaneous point gains and losses on human persistence

de Méo Luiz et al. (2025) ran four small lab tests with neurotypical adults. Everyone pressed a button to earn points on a VI schedule.

Half the group could also lose points after some responses. The team tracked response rate and how long each person kept pressing when the points stopped.

Point losses cut response rate, but they did not change how long people kept pressing later. Persistence stayed the same whether losses were on or off.

Gains-only and gains-plus-losses ended with nearly identical durability.

Charlop et al. (1985) saw the same pattern: losses trimmed rate but did not wreck work output in a three-person token game. Both studies show losses suppress moment-to-moment responding yet leave the staying power of the behavior intact.

Fuller et al. (2018) took the opposite angle. They raised mastery from 50 % to 90 % and got better maintenance for kids with autism. That result says procedural tweaks can strengthen persistence, but the new paper shows losses are not one of those tweaks.

Kuroda et al. (2020) and Lattal (2020) both report null effects for special reinforcer sizes (jackpots). The 2025 data line up: unusual magnitude events—whether big jackpots or small losses—fail to give extra behavioral punch.

If you are trying to build durable skills, pour energy into rich, steady gains instead of adding penalties. Losses may slow the learner down in the moment, yet they do not make the behavior more fragile later. Keep reinforcement strong and consistent; skip the side of punishment if your goal is long-term maintenance.