Low frequency bi-syllabic wordlists in a South-Indian language, Kannada: development, standardization and validation

Abstract

Objective: The present study aimed to develop, standardize and validate low frequency bi-syllabic wordlists in Kannada, a South-Indian language.

Study design: The study was conducted in three different phases. The Kannada low frequency wordlists were developed in Phase I. The procedure involved collecting bi-syllabic familiar words, recording them, selecting the words dominant in low frequency energy by acoustical (Fast Fourier) transform and statistical means (k-means clustering) then generating equivalent wordlists using psychometric function. In Phase II, all the wordlists developed were standardized through estimation of speech identification scores in 100 individuals with normal hearing and through re-verification of equivalence of wordlists’ difficulty level by obtaining psychometric function. Finally, during Phase III, lists developed were evaluated for usefulness by administering them along with conventional phonemically-balanced Kannada wordlist on 10 individuals with cochlear hearing loss having rising audiometric configurations (i.e. more loss at lower frequencies).

Results: Phase I resulted in development of seven psychometrically equivalent wordlists. Speech identification scores on 100 individuals with normal hearing showed mean scores greater than 95% for all the lists at 40 dB SL. No statistical difference was noted across wordlists. Further, individuals with rising cochlear hearing loss (RCHL) performed significantly poorer when compared to normal hearing counterparts across wordlists except low frequency wordlist 4 and phonemically balanced wordlist.

Conclusions: The study utilized a unique procedure for the development of wordlists which can serve as guidelines for further research. The study has resulted in standardization (along with generation of normative data) and successful validation of the lists (except list 4) developed on a clinical population, i.e. individuals with RCHL. Given the lack of availability and the current clinical/research need of such test materials, the wordlists generated from this study can be useful.

Keywords:

• Wordlists
• speech identification
• low frequency
• Kannada
• development

Acknowledgements

We express our sincere gratitude to all the research participants involved in the study. We also thank the All India Institute of Speech and Hearing for its support in data collection.

Disclosure statement

The authors report no conflicts of interest.