Development and content validation of language-free stimuli for evaluating speech intelligibility

Abstract Background Speech intelligibility is the degree to which the speaker’s intended message is recovered by the listener. There are several measures available for speech intelligibility assessment. Multiple studies have examined speech intelligibility assessments across different languages and populations. However, there remains a lack of specific speech intelligibility tests for Indian languages. The present study aims at the development of test stimuli and content validation of language-free word stimuli to assess speech intelligibility. Method The study was conducted in four phases which includes the following: enumerating consonants available in the Dravidian language family and content validation; development and content validation of language-free word stimuli to assess speech intelligibility; content validation of audio-recorded language-free word stimuli, and development and content validation of rating scale and administration of test stimuli in neurotypical individuals . Results Initially, a total of 17,756 words using combination generation in Excel, which was narrowed down to 4000 trisyllabic words with the vowel /a/. Further refinement resulted in 348 trisyllabic words based on the frequency of phonemes in Dravidian languages, considering all consonant positions. Content validation was conducted using the Delphi technique. The final selection of trisyllabic words was based on the content validation ratio (CVR) and content validation index (CVI) scores, resulting in 119 relevant words. The recording of those words was carried out by four speakers (two male and two female). Those recordings were validated by two speech-language pathologists using a ranking system. Based on the ranking, the investigator selected a male speaker who had the best clarity as a source of recording. The investigator administered the prerecorded trisyllabic word stimulus to the participants (30 neurotypical individuals; the age range of 19–65 years) bilaterally. The investigator asked the participant to repeat those words, whatever they heard through the headphones, and those responses were transcribed and analysed by the primary investigator. Out of a total of 119 words, 45 were eliminated from the list because they were (neurotypical individuals) produced incorrectly. The final word list contains 74 trisyllabic words. Conclusion The language-free stimuli can be used for the speech intelligibility assessment of native Dravidian language speakers with dysarthria, neurological conditions like Parkinson’s disease, bulbar palsy, post-glossectomy patients, etc. Future studies can focus on the clinical validation of the stimuli among disorder populations such as dysarthria.