Research Foundations

Key Finding

MFA-SAT reached 86% accuracy on child speech (ages 3-6), the only system approaching human interrater agreement.

VoxKit Implementation

VoxKit defaults to MFA while supporting alternative engines for comparative research.

Critical Consideration

MFA-SAT was trained on adult speech; researchers should validate performance on their specific datasets.

Key Finding

Vowels showed 83% accuracy across systems. Fricative accuracy improved significantly with child age (OR = 1.29/year).

VoxKit Implementation

VoxKit tracks alignment metadata and speaker age, enabling age-stratified accuracy analysis.

Critical Consideration

These patterns emerged from elicited single-word productions and may not generalize to spontaneous speech.

Key Finding

Most clinical speech datasets contain only minutes to hours of audio with uncertain labels, leading to poor generalization.

VoxKit Implementation

VoxKit enables comprehensive metadata tracking and versioned provenance for transparent reporting.

Critical Consideration

VoxKit facilitates rigorous documentation but cannot solve fundamental overfitting—researchers must apply appropriate validation.

Key Finding

Low-dimensional, clinically grounded measures (e.g., hypernasality, articulatory precision) outperform opaque embeddings.

VoxKit Implementation

VoxKit's analyzer architecture supports custom, interpretable feature extraction from alignment outputs.

Critical Consideration

Alignment errors can propagate to downstream measures. Researchers must validate that phonetic boundaries are reliable for their specific analyses.