Objective. This study aims to evaluate the effectiveness of the Cantonese Perceptual Evaluation of Voice (CanPEV) training platform in improving rating accuracy in speech-language pathology students.

Methods/Design: This study is a prospective, single-group, pretraining and posttraining study design. Forty-two speech-language pathology students were recruited for this study. Auditory-perceptual voice training was provided using the CanPEV training platform over 12 weeks. The training platform included three components: (i) familiarization with vocal parameters, (ii) single vocal parameter training, and (iii) multiple vocal parameters training. Students were instructed to utilize the training platform at their own pace and based on their learning needs. All audio samples consisted of sustained vowels, a passage reading, and a conversational speech sample. Students judged voice samples based on ten vocal parameters in the CanPEV scale.

Results: By applying clinician level between the pretraining and the posttraining tests, significant increases in rating accuracy were observed for overall severity, roughness, strain, phonation breaks, glottal fry, tremor, pitch, and loudness (all P .05).

Conclusions: These results supported that the CanPEV training platform is effective in improving rating accuracy in students. The training platform utilizes natural voice samples, an immediate feedback paradigm, and learning at their own pace, which is believed to have facilitated the development of students’ internal standards of vocal parameters.