BACKGROUND: The purpose of this study was first to explore the cross-linguistic robustness of the Acoustic Voice Quality Index (i.e., AVQI) when administered to German continuous speech segments instead of Dutch sentences. The second aim was to define a normative AVQI threshold for distinguishing between normophonia and dysphonia in German speakers. METHODS: Sixty-one German subjects with diverse voice disorders were asked to sustain the vowel [a] and to read aloud a common text. A 3-s mid-vowel segment and the first two sentences of the text "The northwind and the sun" were edited, concatenated and analyzed according to methods described elsewhere. The voice recordings from all 61 participants were (1) auditory-perceptually rated with 'H' from the RBH system by five experienced clinicians and (2) acoustically analyzed to yield the AVQI scores. RESULTS: First, a reasonable correlation was found between the AVQI and H (i.e., r(s) = 0.79). Second, the AVQI revealed an acceptable diagnostic differentiation between normal and dysphonic voices (i.e., AUC = 0.888). These results on German material accord with the results from previous studies on Dutch material. Furthermore, in the German version, the AVQI's cutoff score of 2.70 is accompanied by sensitivity = 79% and specificity = 92%. This indicates minimal normative deviation from the AVQI's cutoff score of 2.95 in Dutch. CONCLUSION: The present and the previous studies yielded almost identical results, denoting the AVQI's cross-linguistic robustness and its feasibility to clinically measure voice quality in German.
BACKGROUND: The purpose of this study was first to explore the cross-linguistic robustness of the Acoustic Voice Quality Index (i.e., AVQI) when administered to German continuous speech segments instead of Dutch sentences. The second aim was to define a normative AVQI threshold for distinguishing between normophonia and dysphonia in German speakers. METHODS: Sixty-one German subjects with diverse voice disorders were asked to sustain the vowel [a] and to read aloud a common text. A 3-s mid-vowel segment and the first two sentences of the text "The northwind and the sun" were edited, concatenated and analyzed according to methods described elsewhere. The voice recordings from all 61 participants were (1) auditory-perceptually rated with 'H' from the RBH system by five experienced clinicians and (2) acoustically analyzed to yield the AVQI scores. RESULTS: First, a reasonable correlation was found between the AVQI and H (i.e., r(s) = 0.79). Second, the AVQI revealed an acceptable diagnostic differentiation between normal and dysphonic voices (i.e., AUC = 0.888). These results on German material accord with the results from previous studies on Dutch material. Furthermore, in the German version, the AVQI's cutoff score of 2.70 is accompanied by sensitivity = 79% and specificity = 92%. This indicates minimal normative deviation from the AVQI's cutoff score of 2.95 in Dutch. CONCLUSION: The present and the previous studies yielded almost identical results, denoting the AVQI's cross-linguistic robustness and its feasibility to clinically measure voice quality in German.