Rethinking Prosody Production in Autism: Nuanced Insights From Individual Differences and Network Analysis Approaches.

Researchers studied how 66 autistic and non-autistic children speak, focusing on voice patterns like pitch and speed. While autistic children showed some differences (higher pitch variation, slower speech), when researchers looked at individual patterns, they found three different groups that included both autistic and non-autistic children. This suggests there's no single 'autistic voice' - instead, autistic children have diverse speaking patterns, just like everyone else.

This study examined prosodic features (speech rhythm, pitch, timing) in 66 autistic and non-autistic school-age children and adolescents using narrative generation tasks. Between-group analyses found autistic participants had greater pitch range and variation, plus slower speech and articulation rates. However, network analysis revealed three distinct communities of participants clustered by prosodic features that did not effectively distinguish between autistic and non-autistic groups. The findings suggest prosody in autism may be 'different in different ways' rather than following a single pattern, highlighting significant individual variation within the autistic population and questioning traditional group-difference approaches.

Findings suggest moving away from assumptions about a single 'autistic voice' toward individualized prosodic assessment. Speech therapy should consider individual language skills when addressing prosodic differences. The heterogeneity in prosodic features supports personalized intervention approaches rather than one-size-fits-all treatments.

Sample size of 66 participants may limit generalizability. The study focused on school-age children, so findings may not apply to other age groups. The abstract doesn't specify methodological details about acoustic analysis procedures or potential confounding variables.

Prosodic differences between autistic and non-autistic individuals are recognized, but there is a lack of consensus on the specific prosodic features that characterize the "autistic voice" due to widespread heterogeneity and mixed findings. This study seeks to build further understanding of the nuances of prosody in autism through individual differences and network analyses. Acoustic analyses were conducted from 66 school-age autistic and non-autistic children and adolescents' narrative generation. Between-groups analyses of pitch- and timing-related prosodic features were conducted, followed by within-group analyses investigating associations between prosodic features and individual differences in overall language skills.

Thereafter, established network analysis methods were adopted to detect the communities of participants based on similar prosodic features. Initial between-groups analyses revealed greater pitch range and variation among autistic compared to non-autistic participants, as well as slower speech and articulation rates, although subsequent analyses revealed that speech and articulation rates were associated with overall language skills. Similar to Weed et al. (2024), the community detection algorithm identified three communities of participants clustered by prosodic features (pitch variation, speech and articulation rates, jitter), with various proportions of autistic participants in each community that did not effectively distinguish between autistic and non-autistic participants. Although between-groups differences consistent with similar previous literature have been indicated, community detection analyses further support the notion that prosody in autism may be "different in different ways." This work highlights the importance of moving beyond group-difference approaches in uncovering nuances to individual differences in prosody via within-group and data-driven analysis approaches. https://doi.org/10.23641/asha.31011862.