Different criteria affect prevalence of relapse of behavior targeted for treatment

Podlesnik et al. (2025) looked at two ways to score relapse. One rule counts a relapse only if the top five worst moments are above baseline. The other rule counts relapse if the average of just two moments is above baseline.

They ran the same clinical data through both rules. The clients had intellectual or developmental disabilities. The goal was to see which rule catches more returning problem behavior.

The stricter max-of-5 rule said relapse was rare. The looser mean-of-2 rule said relapse was common. Using the tougher rule hides treatment failure.

Williams et al. (2023) also watched how we measure relapse. They tracked caregiver mistakes moment-by-moment and saw every caregiver slip back. Their continuous method backs up the idea that tight rules miss real relapse.

Jarrold et al. (1994) showed that behavior can return with a new function. If you only count the old form, you will still miss failure. Their warning pairs well with the new math: loose criteria catch both old and new problems.

King et al. (2025) showed that even smart relapse plans can let a few responses slip through. Their data fit the mean-of-2 picture: a couple of strong responses matter.

If you use the max-of-5 rule, you may discharge a client too early. Switch to mean-of-2 and you will spot back-slide sooner. Add booster sessions when you see any two data points rise. This small math tweak can save weeks of re-doing treatment.