Speech disfluencies in typically developing Finnish-speaking children - preliminary results

Area of Science:

• Speech-language pathology
• Developmental linguistics
• Childhood communication disorders

Background:

• Understanding typical speech disfluency development is crucial for identifying stuttering.
• Previous research on disfluency norms has varied across languages and age groups.
• Finnish language presents unique phonetic and prosodic features that may influence disfluency patterns.

Purpose of the Study:

• To investigate the frequency and types of speech disfluencies in typically fluent Finnish-speaking children across different age bands.
• To evaluate the applicability of existing criteria for distinguishing normal disfluency from stuttering in young children.
• To determine if disfluency measurement guidelines need language-specific adaptations.

Main Methods:

• Analysis of speech samples from 54 typically fluent Finnish-speaking children (ages 2-4 and 6-9 years).
• Data collection through semi-structured conversational interviews.
• Quantification of disfluencies per 100 words and per 100 syllables, categorizing disfluency types.

Main Results:

• No significant differences in total disfluency frequency or types were found among 2- to 4-year-olds.
• Significant between-group differences in total disfluency frequency and types were observed in 6- to 9-year-olds.
• The criterion of <3 stuttering-like disfluencies (SLD) per 100 syllables effectively distinguished normal fluency across all age groups, unlike the <3 SLD per 100 words criterion.

Conclusions:

• The frequency and types of disfluencies in typically fluent Finnish children show developmental changes between early childhood and later childhood.
• The metric of stuttering-like disfluencies per 100 syllables appears to be a more robust indicator of normal fluency across different age groups compared to measurement per 100 words.
• Preliminary findings suggest the need for language-specific guidelines in defining stuttering based on disfluency counts.