Equivalence of the Matrix Sentence Test in Indian-English in the Presence of Noise

Abstract

The objective of the study was to evaluate the equivalence of sentence lists of the 'Matrix Sentence Test in Indian- English' developed by Bhattarai and Yathiraj (2015) in the presence of noise as well as determine the difference in performance between children and young adults.A purposive sampling technique was used to select the participants. Initially, a pilot investigation was done to determine the appropriate signal-to-noise ratio to establish SNR-50; a pilot study was conducted on 5 normal hearing children and 5 normal hearing young adults. The sentences were presented at a constant level, equivalent to normal conversation (45 dB HL), and an 8-talker babble was varied to establish SNR-50. SNR-50 was obtained at 0 dB SNR for the children and -5 dB SNR for the young adults. Using these values, SNR-50 was evaluated on 60 participants. These participants were divided into four age groups with 10 individuals each in the younger three age groups [7 year olds (7.1 to 8 years), 8 year olds (8.1 to 9 years), 9 year olds (9.1 to 10 years)] and 30 young adults aged 18 to 25 years. No significant effect of age was found on the sentence identification scores in the presence of speech babble for 48 lists of the 50 lists of the 'Matrix Sentence Test in Indian-English'. The lists, 23 and 33 that were found to vary across ages were eliminated for further analyses. Among the 48 lists, it was found that 37 lists were equivalent and 11 lists were unequal in the presence of noise. The 48 sentence lists that had no age effect can be used to test sentence identification scores for both children and adults in the presence of noise. Among these, 37 lists can be used interchangeably. It is suggested to regroup the 11 unequal lists after eliminating sentences with specific words that yielded poor scores.