Differential reinforcement as treatment for behavior disorders: procedural and functional variations.

Demello et al. (1992) wrote a big-picture review of differential-reinforcement treatments.

They looked at how different ways of delivering DR match, or miss, the reason the problem behavior happens.

The paper warns that weak or skipped functional assessments leave clinicians guessing.

The review says DR works best when the reinforcer you give is the same one the child gets from the problem behavior.

When the reinforcer is wrong, success drops.

The authors call for tighter, faster ways to find function before treatment starts.

Perez et al. (2015) later pooled data and showed noncontingent reinforcement cuts problem behavior by a very large margin, but only when the reinforcer is functional.

Fahmie et al. (2013) ran an experiment and found DRO fails as a control condition if negative reinforcement keeps the behavior alive, backing the review’s warning.

Fisher et al. (2016) seems to clash: their traditional FA found functions that the newer IISCA missed. The issue is assessment format, not DR itself; the review would still say pick the test that truly shows function, then match DR to it.

Verriden et al. (2019) extended the story: when DR plus NCR did nothing for automatically reinforced behavior, adding a brief punisher test gave big gains.

You can save hours of failed treatment by running any quick, valid FA first. Pick the reinforcer that matches the identified function, then choose your DR type. If the behavior is automatic and DR stalls, test tiny punisher additions instead of abandoning the case.