Modelling category goodness judgments in children with residual sound errors

Area of Science:

• Linguistics
• Speech Science
• Developmental Psychology

Background:

• Category goodness judgments are crucial for understanding speech perception.
• The acoustic properties of the /r/ sound, specifically the third formant (F3), are key perceptual cues.
• Children with residual speech errors (RSE) may exhibit different speech perception patterns compared to typically developing (TD) peers.

Purpose of the Study:

• To investigate how adults and children, with and without RSE, make category goodness judgments of the /r/ sound.
• To determine the influence of normalized third formant frequency (F3) and vowel context on these judgments.
• To explore the relationship between category goodness judgments and /r/ production abilities in children with RSE.

Main Methods:

• Adults and children (ages 7-16) with and without RSE provided category goodness judgments on natural /r/ speech stimuli.
• Acoustic analysis normalized the third formant (F3) frequency of /r/ using two methods.
• Logistic mixed-effect modeling analyzed the relationship between listener judgments, F3, vowel context, and group status.

Main Results:

• Adult judgments showed a significant interaction with normalized F3 compared to both child groups (p < 0.001).
• Children with RSE did not significantly differ from TD children in their /r/ judgments.
• All listeners favored /r/ in front-vowel contexts, and normalized F3 predicted judgments, but children's judgments were not adult-like.

Conclusions:

• Normalized /r/ F3 significantly predicts category goodness judgments in adults and children, though children's perception differs from adults.
• Category goodness judgments do not appear to correlate with /r/ production abilities in children with RSE.
• Findings may inform clinical practices involving natural speech category judgments, particularly for recorded productions.