Development of High Frequency Speech Identification Test in Manipuri Language

Abstract

The present study aimed at developing and administering high frequency speech identification test in Manipuri language. The study consisted of two stages. In stage I, monosyllabic words, which majorly contained k/, /kh /, /h/, /s/, /p/, /ph/, /t/, /th/, /t /I & /f/ consonants and /i/,/e/ & /ei/ vowels, were selected from different sources. Long-term average speech spectrum was done on the selected words, to confirm that the selected words had high frequency spectral energy. Words with peak frequency of 2 kHz or above, and which had energy present even in the frequencies above the peak frequency were selected. After this, familiarity check was done. Only familiar and most familiar words were selected and two sub tests of monosyllabic high frequency words were developed using those words. The developed materials were administered at three presentation Levels [20, 40 and 60 dB SL (Ref: SRT)] to 20 normal hearing individuals at each level. SIS was obtained using both the sub tests. The results showed a significant difference between the three Levels, i. e., with increase in the levels, the SIS increased, with almost 100% at 40 dB SL. Comparison across the two sub tests showed that any of the two sub tests can be used to obtain high frequency speech identification scores. Thus, from the above findings, it may be concluded that this test has got normal performance-intensity function and has List equality. However, the clinical utility of the test has to be assessed by administering it on clinical population with different degrees of hearing impairment.