An evaluation of interrupted and uninterrupted measurement of vocal stereotypy on perceived treatment outcomes.
Count every interval—skipping post-interruption windows makes RIRD look stronger than it is.
01Research in Context
What this study did
The team watched three boys with autism during free-play. Each child got two treatments for loud repetitive sounds: RIRD where the therapist interrupted and redirected every stereotypic sound, and NCR where the child heard favorite music no matter what.
Sessions were 10 minutes long. The researchers scored vocal stereotypy two ways: counting every 10-second interval (uninterrupted) or skipping intervals right after an interruption (interrupted).
What they found
When they skipped the post-interruption intervals, RIRD looked like a big winner—stereotypy dropped sharply. The uninterrupted count told a different story; RIRD barely beat NCR.
NCR looked the same no matter how they scored it. The way you slice the data, not the treatment itself, created the hero effect for RIRD.
How this fits with other research
Gibbs et al. (2018) later paired the same RIRD procedure with noncontingent music and saw faster gains. Their positive results make sense only if you count every interval; otherwise the "boost" could just be the old scoring trick.
Cividini-Motta et al. (2019) added a DRA component to RIRD and still found weak effects on appropriate play. Their mixed findings line up with the uninterrupted counts here—RIRD cuts stereotypy but don’t expect new skills without extra teaching.
Rojahn et al. (1987) warned that short gaps between conditions can fake treatment effects in alternating designs. The new paper shows a second trap: skipping intervals can inflate success even when the design is sound.
Why it matters
Before you praise RIRD, score every 10-second window. If your data sheet auto-skips post-redirection intervals, you may celebrate too early. Use uninterrupted measurement and you will see whether RIRD, NCR, or music really wins for that child.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Open your timing sheet and ensure you mark stereotypy in every 10-second block, even right after you redirect.
02At a glance
03Original abstract
The type of procedure used to measure a target behavior may directly influence the perceived treatment outcomes. In the present study, we examined the influence of different data-analysis procedures on the outcomes of two commonly used treatments on the vocal stereotypy of 2 children with an autism spectrum disorder. In Study 1, we compared an interrupted and uninterrupted data-analysis procedure to measure vocal stereotypy during the implementation of response interruption and redirection (RIRD). The results showed that the interrupted data-analysis procedure overestimated the effectiveness of RIRD. In Study 2, we examined the influence of different data-analysis procedures on the interpretation of the relative effects of 2 different treatments for vocal stereotypy. Specifically, we compared interrupted and uninterrupted data-analysis procedures during the implementation of RIRD and noncontingent reinforcement (NCR) as a treatment for vocal stereotypy. The results showed that, as in Study 1, the interrupted data-analysis procedure overestimated the effectiveness of RIRD; however, this effect was not apparent with NCR. These findings suggest that different types of data analysis can influence the perceived success of a treatment.
Journal of applied behavior analysis, 2014 · doi:10.1002/jaba.118