Further Translational Evaluations of Efficacy and Preference for Isolated and Synthesized Contingency Procedures

Auten et al. (2025) compared two ways to build a DRA plan after a functional analysis. One plan used an isolated contingency. The other used a synthesized contingency. They tested both plans with the same children in an alternating-treatments design.

The team tracked problem behavior and appropriate communication during each plan. After each session they asked the child which plan felt better. The goal was to see if one plan worked faster and which one the child liked.

Both the isolated and the synthesized DRA cut problem behavior and boosted communication. The gains looked the same across the two plans.

The twist: each child liked one plan more than the other. The preference was not predictable from the FA results. You have to ask the kid.

Nevin et al. (2016) showed that lean signaled DRA keeps relapse low after treatment ends. Auten’s work adds that either isolated or synthesized DRA can be that lean signaled plan, so you can pick the one the child prefers without risking more relapse.

Jackson et al. (2026) later showed that fading the therapy context back in cuts renewal in half. Their DRA plans grew from the same translational line Auten started, showing the field is moving from ‘which plan works’ to ‘how do we make it stick.’

Dowdy et al. (2020) took DRA without extinction into a public pool and still saw big drops in problem behavior. That real-world success backs Auten’s lab finding: DRA is robust across settings, so letting the child choose the contingency style is not a luxury—it’s doable in the community.

You no longer need to agonize over isolated versus synthesized DRA. Both work. Run a brief comparison, let the client vote with their feet, and move on. The child gets a voice, you get the same behavior drop, and future relapse is still covered by lean signaled delivery. Next time you write a behavior plan, add a 5-minute preference check between the two DRA setups—then honor the winner.