Development and evaluation of the digit triplet and auditory-visual matrix sentence tests in Malay
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
The objective of this study is to develop and evaluate versions of the digit triplet (MDTT) and matrix sentence tests in Malay (MMST-AV). The development of both tests involved the selection, recording and normalisation with level adjustments of speech stimuli in the form of digits and words. The MDTT was developed for headphone and telephone applications in test specific noise (TSN) and spectrotemporal gap noise (STG). The MMST-AV was developed for headphone application in TSN and 6-talker babble noise (BN). To allow for auditory-visual (AV) mode of assessment of the MMST-AV, a visual component was added which required additional studies to investigate the optimal method of normalisation and the refinement of video samples. Both tests were evaluated auditorily for list equivalency in normal hearing listeners and were further validated in a group of listeners with varying hearing levels. Additionally, the evaluation of the MMST-AV included the investigation of the training effects. Eight lists of 27 digit triplets and 15 lists of 30 sentences were evaluated monaurally in a closed-set response format for the MDTT and MMST-AV, respectively. A total of 166 normal hearing and 26 hearing impaired participants were recruited for this study. For the MDTT, evaluation in fixed SNRs resulted in a mean speech reception threshold (SRT) of -11.3 ± 0.34 dB SNR for headphone application in TSN; -11.9 ± 0.4 dB SNR for headphone application in STG; -10.24 ± 0.1 dB SNR for telephone application in TSN and -10.8 ± 0.3 dB SNR for telephone application in STG. The mean SRT and slope normative reference for the MMST-AV were -10.1 ± 0.2 dB SNR and 14.9 ± 1.2 %/dB, respectively in TSN whereas in BN, mean SRT and slope were –6.4 ± 0.2 dB SNR and 12.2 ± 0.7 %/dB, respectively. A significant training effect of 1.4 dB was observed for the first two consecutive measurements in the TSN and 0.8 dB in BN. Evaluation in listeners with varying hearing levels in the MDTT revealed test sensitivities and specificities of more than 85% in all four test conditions. Performances of normal and hearing impaired groups were found to be equal in the AV and visual-only mode of testing after the effect of participant’s age was controlled. In conclusion, both the MDTT and MMST-AV showed good agreement between the SRTs and slope to other versions of the tests and are suitable for repeated measurements.