A Winkler1, I Holube2. 1. Institut für Hörtechnik und Audiologie, Jade Hochschule und Exzellenzcluster "Hearing4All", Oldenburg, 26121, Oldenburg, Deutschland. Alexandra.Winkler@jade-hs.de. 2. Institut für Hörtechnik und Audiologie, Jade Hochschule und Exzellenzcluster "Hearing4All", Oldenburg, 26121, Oldenburg, Deutschland.
Abstract
BACKGROUND: ISO 8253-3 describes the requirements for speech tests used in audiology. One aspect is the test-retest reliability, which was here examined for the Freiburg monosyllabic speech test. METHODS: Speech recognition scores were examined for 30 listeners with normal hearing ability. All 20 lists of the Freiburg monosyllabic speech test were tested at four different presentation levels. The retest was carried out 7-9 months later. RESULTS: The test-retest reliability of the Freiburg monosyllabic speech test in this study is between 13 percentage points for a speech recognition score of 95 %, and 24 percentage points for a speech recognition score of 39 %. The measurement accuracy can be estimated using binomial distribution. CONCLUSION: The reproducibility of the results is limited by the small number of test items (20 words) in each list. This does not fulfill the requirements for speech tests in diagnostics and rehabilitation. An improvement of the test-retest reliability can be achieved using several test lists.
BACKGROUND: ISO 8253-3 describes the requirements for speech tests used in audiology. One aspect is the test-retest reliability, which was here examined for the Freiburg monosyllabic speech test. METHODS: Speech recognition scores were examined for 30 listeners with normal hearing ability. All 20 lists of the Freiburg monosyllabic speech test were tested at four different presentation levels. The retest was carried out 7-9 months later. RESULTS: The test-retest reliability of the Freiburg monosyllabic speech test in this study is between 13 percentage points for a speech recognition score of 95 %, and 24 percentage points for a speech recognition score of 39 %. The measurement accuracy can be estimated using binomial distribution. CONCLUSION: The reproducibility of the results is limited by the small number of test items (20 words) in each list. This does not fulfill the requirements for speech tests in diagnostics and rehabilitation. An improvement of the test-retest reliability can be achieved using several test lists.
Authors: Piotr H Skarzynski; Anna Ratuszniak; Kamila Osinska; Magdalena Koziel; Bartlomiej Krol; Katarzyna B Cywka; Henryk Skarzynski Journal: Otol Neurotol Date: 2019-08 Impact factor: 2.311