Examining the Effects of Jasper and Enhanced Milieu Teaching on Repetitive Behaviors and Scripted Language

Background: Restricted and repetitive behaviors, interests, and activities (RRBs) are one of two core diagnostic features of autism spectrum disorder (American Psychiatric Association, 2013). RRBs define a broad category of behaviors, including motor stereotypies, unusual sensory preoccupations, and echolalia. RRBs have been shown to interfere with child learning (Koegel & Covert, 1972) and have been the focus of a broad range of behavioral interventions (Rapp & Vollmer, 2005). The current study examined the extent to which RRBs change over the course of an intervention targeting social attention, play and spoken language, and the relationship between scripted language and social communicative utterances.

Objectives: 1) What is the relationship between parent-reported RRBs, the repetitive behavior score on the ADOS, and observed measures of RRBs? 2) Do these measures change over the course of intervention? 3) Over the course of intervention, does observed repetitive behavior decrease as language increases?

Methods: Twenty minimally verbal (less than 20 words) children diagnosed with ASD between the ages of 4.5 and 8 years old participated (16 males, mean age 6.39). The intervention used the strategies associated with JASPER/Enhanced Milieu Teaching (Kasari et al, submitted) to teach joint attention, nonverbal and verbal requests, and comments. Data were selected from the pre and post intervention assessments, approximately 3 months apart. The repetitive behavior subscale of the ADOS was measured pre-intervention, and parent-reported Repetitive Behavior Scale (RBS), social communicative utterances (SCU), and scripted language were measured pre- and post-intervention.  

Results: Results show that there was not a significant correlation between ADOS repetitive behavior subscale and pretest RBS scores (r=.202, p=.436), or between ADOS repetitive behavior subscale and scripted language on the language sample at pretest (r=.123, p=.606). Over the course of intervention, while total social communicative utterance did show significant changes (t = -2.3918, p = 0.02727), RBS did not show significant changes from pretest to posttest (t = 0.414, p = 0.684). RBS remained low pre- and post-test (pre test mean=30.11, sd=20.27, post test mean=34.15, sd=24.23). Decreases in scripted language only approached significance (t = 1.7147, p= 0.1027). Changes in social-communicative utterances were not significantly related to changes in scripted language over the course of intervention (r=.276 p=.239).

Conclusions: There are several possible explanations for the nonsignficant findings. First parents rated their children as having low levels of RBS although observations during the ADOS suggested moderately high levels of repetitive behavior.  Second, some children had extremely low rates of any language behavior during the language sample session, including both social communicative utterances and scripted language.  These factors, combined with the small sample size, may have affected the outcomes.  To further examine the functional relationship between behavior and language during the sessions, we are coding all repetitive behaviors, during the first two intervention sessions compared to the final two intervention sessions. We hypothesize that the inclusion of motor stereotypies as well as verbal scripts will provide a more representative data set to answer to the question of correlation between changes in observed RRBs and language over the course of intervention.