Correspondence between preference for and efficacy of behavioral interventions: A systematic review
Clients usually like the intervention that works best, especially if it’s on a screen.
01Research in Context
What this study did
Goldman et al. (2025) pulled every paper that tested both how much clients liked an intervention and how well it worked.
They found 457 single cases and asked one simple question: does the favorite match the most effective?
What they found
Three out of four times the intervention people liked best was also the one that cut problem behavior the most.
Computer or tablet tasks won on both counts—they stayed at the top for preference and success.
How this fits with other research
Campbell (2003) and Heyvaert et al. (2014) already showed that doing a functional analysis first gives stronger behavior reduction. Goldman’s data say the client will usually pick that same FA-based plan, so you no longer have to choose between “what works” and “what they’ll accept.”
Kok et al. (2026) warned that gains often fade after sessions stop. Goldman’s review hints at a fix: when clients prefer the procedure, they may keep using it, helping effects last.
Grynszpan et al. (2014) found a medium skill boost from tech tools. Goldman shows these tools are also the most liked, raising the odds that families will stick with them.
Why it matters
You can stop guessing. Run a quick preference assessment and an efficacy probe; if one intervention wins both, you have a green light. Start with tablet or computer programs when they fit the goal—they rarely lose.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Put the top two interventions in a brief client-choice test; pick the one that wins both preference and initial data.
02At a glance
03Original abstract
Understanding of the correspondence between intervention preference and efficacy is limited. We systematically reviewed 112 articles (457 cases) evaluating efficacy of and preference for behavioral interventions. We analyzed the percentage of cases for whom interventions were preferred and efficacious across broad (e.g., behavior reduction, performance, skill acquisition) and specific (e.g., noncontingent reinforcement, video modeling) intervention types. Authors reported one preferred intervention for most cases. Regarding efficacy, authors reported about half of cases as having one efficacious intervention and the other half having multiple equally efficacious interventions. The same intervention was preferred and efficacious for 74% of cases for whom authors reported one preferred and one efficacious intervention. Several specific interventions were generally preferred and efficacious across cases (e.g., digital stimuli, computer-based instruction, accumulated reinforcement, contingent reinforcement). We discuss clinical recommendations, the importance of assessing preference, and the need for research in developing protocols for assessing intervention preference.
Journal of Applied Behavior Analysis, 2025 · doi:10.1002/jaba.2924