This DSI-SRP fellowship funded Yunah Song to work in the laboratory of Dr. Miriam Lense in the Department of Otolaryngology during the summer of 2022. Yunah is a junior with majors in Cognitive Studies and Child Development.
Timing is an integral component of successful parent-child social interactions for both autistic children (ASD) and children who are typically developing (TD). Timing in parent child interactions involves the rhythm and frequency of individuals’ vocalizations and the relative timing between the two partners’ vocalizations. The Music Cognition Lab has previously observed differences in the acoustic hierarchical temporal structure, a measure of the nested clustering of acoustic events across multiple timescales (i.e., how phonemes are nested in syllables, which are nested in words, which are nested in phrases, which are nested in sentences, and which are then nested within turns between individuals and interaction patterns), of parent-child interactions between TD and ASD parent-child dyads (Boorom, Alviar et al., 2022). Specifically, in parent-child dyads of TD infants and toddlers, acoustic hierarchical temporal structure decreases with increasing child language skills, reflecting vocal interactions that become increasingly flexible. In contrast, parent-child dyads of autistic toddlers exhibit greater hierarchical temporal acoustic structure compared to TD dyads of similar developmental or expressive language levels, potentially reflecting less flexible temporal structure and differences in social reciprocity in ASD and TD parent-child dyads.
In the current study, Yunah and Dr. Lense provided a parent coaching intervention that taught parents of children with ASD how to more effectively engage with and support their social interactions with their child. They examined changes in the dynamics of the parent-child interaction by assessing changes in the dyads’ acoustic temporal structure over the course of the parenting intervention. Parent-child dyads were audio-video recorded in 10-minute free play interactions as part of their weekly participation in the 10-week intervention. The acoustic temporal structure of the interactions were measured using a novel and automated metric of hierarchical temporal clustering, Allan Factor variance, which was applied to the audio of the play interactions. The Allan Factor analysis involves extracting acoustic events within specific time windows and then quantifying the nested variability of acoustic events across multiple timescales. They calculated the slope of the Allan Factor function for each recording to measure the multiscale temporal structure of that free play session.
Preliminary exploratory analyses indicate inter- and intra-individual differences in hierarchical temporal structure as dyads participate in the parenting intervention. It is speculated that the specific strategies taught during the intervention may influence the dyads’ acoustic temporal structure. For example, during the session where parents are taught to imitate their child’s play, the parents seem to display increased (i.e. more clustered), hierarchical temporal structure. This may reflect reductions in parent vocalizations as they are focusing on using nonverbal techniques with their child rather than verbal techniques (e.g., like describing their child’s play). Future directions for this research include examining behavioral correlates (e.g., parent-child shared engagement, parent and child frequency of vocalizations) of observed changes in hierarchical temporal structure throughout the intervention. This study was able to use an automated analysis metric that can be applied to real-world social interactions suggesting it may be useful for dynamically assessing social interaction across individuals, contexts, and development.
In addition to receiving support through a DSI-SRP fellowship, this project was supported and facilitated by the DSI Data Science Team through their regular summer workshops and demo sessions.