Remote Microphone Systems for Autistic and Nonautistic Youth: Effects on Audiovisual Task Engagement
Kacie Dunham-Carr,Nisha Mailapur,Bahar Keçeli-Kaysili,Jacob I Feldman,Emily Thompson,Hilary Davis,Anne Marie Tharpe,Erin Picou,Tiffany G Woynaroski
DOI: https://doi.org/10.1097/AUD.0000000000001581
2024-09-23
Abstract:Objectives: A recent study has provided empirical support for the use of remote microphone (RM) systems to improve listening-in-noise performance of autistic youth. It has been proposed that RM system effects might be achieved by boosting engagement in this population. The present study used behavioral coding to test this hypothesis in autistic and nonautistic youth listening in an ecologically valid, noisy environment. Design: We drew on extant data from a recent experimental study in which 56 youth (32 autistic, 24 nonautistic) matched at the group level on age and biological sex completed listening-in-noise tasks wherein they reported their perception of audiovisual syllables, words, sentences, and passages with and without an RM system; conditions were counter-balanced across participants. As previously reported, perceptual accuracy varied with stimulus complexity and overall improved with the RM system, with improvements not significantly different between groups. Video recordings of participants completing listening-in-noise tasks in both conditions were coded via a 5-second, partial-interval coding system by naive coders for (a) engagement in the task (indexed via proportion of intervals in which participants displayed on-task behaviors) and (b) verbal, stimulus-specific protesting in the task (indexed via proportion of intervals in which participants displayed verbal, stimulus-specific protesting behaviors). Examples of on-task behaviors included attending to the screen and completing task activities. Examples of protesting behaviors included complaining about stimuli volume or the inability to hear. Chronological age, autism features, language ability, audiovisual speech integration as measured by psychophysical tasks, tactile responsiveness, and nonverbal intelligence quotient were evaluated as putative predictors and/or moderators of effects on behaviors of interest. Results: In general, participants were highly engaged in the task, and there were few protests, reflecting more than 90% and fewer than 0.5% of coded intervals, respectively. We did not detect any statistically significant effects of group or RM system use on task engagement. Nonautistic youth were engaged in the listening-in-noise task for an average of 97.45% of intervals, whereas autistic youth were engaged in the listening-in-noise task for an average of 94.25% of intervals. In contrast, verbal, stimulus-specific protesting in the listening-in-noise task was significantly reduced, on average, in the RM (0.04% of intervals) versus the No RM (0.2% of intervals) conditions. There were no effects related to group for this behaviorally coded outcome. In addition, select participant characteristics predicted engagement within conditions across participants. Greater language ability and nonverbal intelligence quotient predicted increased engagement when not using an RM system. Increased features of autism and wider temporal binding windows for audiovisual speech predicted reduced engagement while using an RM system, and greater audiovisual integration predicted increased engagement while using an RM system. Conclusions: The results of this study suggest that RM system use reduces verbal, stimulus-specific protesting, which likely reflects difficulty engaging when listening in noise. The present study extends our previous study to provide additional empirical support for RM system use in autistic and nonautistic youth.