Examining the Spontaneous Communication of Minimally Verbal Children with ASD in Supported Versus Unsupported Contexts

Background:   Despite access to early intervention, up to 30% of children diagnosed with autism spectrum disorder (ASD) are classified as nonverbal or minimally verbal at age five. Being non-verbal or minimally verbal can have a significant impact on adaptive skills and social behaviors for children with autism (Anderson et al., 2007). Previous research has shown increases in child spontaneity and frequency of language for this population with interventions that are focused on responsive interactions (Kasari et al., 2014). Further research is needed to identify key behaviors in responsive communication partners to better treat this population.

Objectives:   The purpose of this study is to examine differences in spontaneous communication for minimally verbal children with ASD in supported and unsupported contexts for communication.

Methods: Twenty-five minimally verbal children with ASD participated in the study (males=21, age= 4.5-8.5 years) as a part of an ongoing study comparing the effects of a naturalistic, play-based intervention to a behavioral intervention (AIM-ASD #426-230-0013). Minimally verbal was defined as a participant displaying fewer than 20 words in a 20-minute language sample. We compared the spontaneous language of participants across two contexts during initial screening assessments: unsupported and supported. In the unsupported context, an adult interacted with the participant using a pre-determined set of toys, but did not model language relevant to the child’s actions or expand on child language utterances. In the supported context, the adult interacted with the child using a pre-determined set of toys and was highly responsive to the child’s communication attempts (mean responsiveness above 90%). The adult contingently imitated the child’s actions and modeled language relevant to the child’s actions. All assessments were video recorded and child utterances were transcribed, verified, and coded for spontaneity. Two-sample, within subject t-tests were completed to analyze the number of different spontaneous words (NDW), the number of total spontaneous words (NTW), and the number of spontaneous utterances using spoken language (NSU). 

Results:   No measures of spontaneous language showed significant differences in a supported interaction compared to an unsupported interaction across three measures: NDW (p=.164), NTW (p=0.872), and NSU (p=0.617). Overall, spontaneous language was low across both contexts: the mean NDW was 10.16 in a supported context (range 0-46) and 7.84 in an unsupported context (range 0-26). The mean NTW was 18.72 in a supported context (range 0-128) and 19.2 in an unsupported context (range 0-121). The mean NSU was 12.08 in a supported context (range 0-74) and 10.96 in an unsupported context (range 0-58).

Conclusions:   Minimally verbal children used few spontaneous utterances in both supported and unsupported contexts. The non-significant results of this analysis show that access to a supportive and engaging communication partner does not show an immediate effect on spontaneous language for this population, which is important in considerations of intervention strategies. For minimally verbal children, extended periods of rich intervention involving a supportive partner is necessary for increasing spontaneity of speech. Analyses will be conducted to evaluate specific behaviors of a supportive partner that may have functioned to increase spontaneous use of language.