Speech intelligibility is a crucial factor in everyday communication which can be estimated under controlled conditions using speech intelligibility tests. Besides the application on diagnostic purposes, speech intelligibility tests can be used for evaluating speech intelligibility in realistic rooms like offices, classrooms or auditoriums. Therefore, well standardized speech intelligibility tests and methods which are comparable across languages are needed. The current work presents a evaluation procedure and the preliminary results of normally hearing listeners of the newly developed Italian Matrix Sentence Test for assessing speech reception thresholds (SRT, i.e. signal-to-noise ratio SNR of 50% intelligibility) in noise. The test is composed of a matrix of ten names, verbs, numerals, nouns and adjectives which get randomly combined to create syntactically fixed but semantically unpredictable sentences. The goal of the evaluation measurements was to investigate the equivalence of test lists, examine the training effect and establish the reference data for normally hearing listeners. Speech intelligibility measured at three fixed SNRs showed comparable list-specific intelligibility functions, which resulted in a mean SRT and slope of -7.3 ± 0.2 dB SNR and 13.3 ± 1.2 %/dB, respectively. To investigate the training effect and establish reference data, speech intelligibility was measured adaptively with 6 test lists of 20 sentences using an open- and a closed-set test format. The major training effect of about 1.5 dB was observed for the first two consecutive measurements which was also found for matrix sentence tests in other languages. From the third till the 6th measurement improvements in SRT were rather small with average thresholds of -6.7 ± 0.7 dB SNR and -7.4 ± 0.7 dB SNR for the open- and closed-set test format, respectively. The optimized and evaluated Italian Matrix Sentence Test will be used for speech intelligibility measurements with pupils in classrooms.

Construction and first evaluation of the Italian Matrix Sentence Test for the assessment of speech intelligibility in noise

PRODI, Nicola;VISENTIN, Chiara;
2014

Abstract

Speech intelligibility is a crucial factor in everyday communication which can be estimated under controlled conditions using speech intelligibility tests. Besides the application on diagnostic purposes, speech intelligibility tests can be used for evaluating speech intelligibility in realistic rooms like offices, classrooms or auditoriums. Therefore, well standardized speech intelligibility tests and methods which are comparable across languages are needed. The current work presents a evaluation procedure and the preliminary results of normally hearing listeners of the newly developed Italian Matrix Sentence Test for assessing speech reception thresholds (SRT, i.e. signal-to-noise ratio SNR of 50% intelligibility) in noise. The test is composed of a matrix of ten names, verbs, numerals, nouns and adjectives which get randomly combined to create syntactically fixed but semantically unpredictable sentences. The goal of the evaluation measurements was to investigate the equivalence of test lists, examine the training effect and establish the reference data for normally hearing listeners. Speech intelligibility measured at three fixed SNRs showed comparable list-specific intelligibility functions, which resulted in a mean SRT and slope of -7.3 ± 0.2 dB SNR and 13.3 ± 1.2 %/dB, respectively. To investigate the training effect and establish reference data, speech intelligibility was measured adaptively with 6 test lists of 20 sentences using an open- and a closed-set test format. The major training effect of about 1.5 dB was observed for the first two consecutive measurements which was also found for matrix sentence tests in other languages. From the third till the 6th measurement improvements in SRT were rather small with average thresholds of -6.7 ± 0.7 dB SNR and -7.4 ± 0.7 dB SNR for the open- and closed-set test format, respectively. The optimized and evaluated Italian Matrix Sentence Test will be used for speech intelligibility measurements with pupils in classrooms.
2014
9788361402282
Audition; Computational linguistics; Matrix algebra; Signal to noise ratio; Speech; Speech communication, Closed set; Consecutive measurements; Controlled conditions; Reference data; Speech intelligibility in noise; Speech intelligibility tests; Speech reception threshold; Training effects, Speech intelligibility
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11392/2362953
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 6
  • ???jsp.display-item.citation.isi??? ND
social impact