Reliability Study of GRBAS and CAPE-V Based on Large Samples in Chinese Context

Area of Science:

• Speech and Hearing Sciences
• Otolaryngology
• Clinical Linguistics

Background:

• Accurate voice assessment relies on consistent perceptual evaluation tools.
• The Grade, Roughness, Breathiness, Asthenia, Strain (GRBAS) and Consensus Auditory-Perceptual Evaluation of Voice (CAPE-V) are widely used but their reliability in diverse linguistic contexts requires verification.

Purpose of the Study:

• To assess the intra-rater and inter-rater reliability of GRBAS and CAPE-V in the Chinese population.
• To compare the consistency of these two voice evaluation scales within the Chinese context.

Main Methods:

• 662 voice samples were evaluated by five experts using GRBAS and CAPE-V across two rounds.
• Inter-rater agreement was analyzed using the first round of data, while intra-rater consistency was assessed using 150 randomly selected samples from the second round.
• Statistical analyses included Cohen's Kappa, Fleiss Kappa, intra-class correlation coefficient (ICC), and Spearman correlation.

Main Results:

• GRBAS demonstrated poor inter-rater agreement (Fleiss Kappa: 0.08-0.32) and fair to moderate intra-rater consistency (Cohen's Kappa: 0.22-0.53).
• CAPE-V showed moderate to excellent intra-rater consistency (ICC: 0.62-0.91) and very high inter-rater agreement (ICC: 0.86-0.92).
• Four shared features (overall voice quality, roughness, breathiness, strain) exhibited high consistency between the scales (r_s: 0.90-0.92).

Conclusions:

• CAPE-V offers superior reliability and precision for voice assessment in the Chinese context, making it ideal for research and clinical monitoring.
• Despite GRBAS's limitations in consistency, its historical clinical utility and ease of rapid assessment currently maintain its relevance.
• CAPE-V is positioned to become the preferred tool as its adoption grows, potentially replacing GRBAS in the future.