[Voice handicap index adapted to the singing voice]

The present study is the outcome of 5 former studies aiming at adapting the V.H.I. to singers. We propose in this last version a V.H.I. valid for all types of singers, dysodic or not.

The reliability, validity and internal coherence were evaluated on a population of 192 classical singers and 27 non singers.

The comparison of the mean scores at the test and the retest shows that the results at the retest are significantly inferior in both the functional scale (p= 0.013) and the global score (p= 0.006). The test/re-test stability is good (ICC: Functional scale (F): 0.869, Emotional scale (E): 0.846, Physical scale (P): 0.85, Global score (G): 0.878/ Spearman's rho F: 0.855, E: 0.834, P: 0.858, G: 0.886). The internal coherence of each sub scale is good (Cronbach's alpha: > 0.80). The correlation between the 3 subscales is quite high (FvsE: 0,613 /EvsP: 0,737/ FvsP: 0,693) which means that each scale measures a specific dimension but that they are still homogeneous. The test/re-test difference between the singers and the controls are not significant but the control group has a greater variability in their answers (Levene test: F: p<0.001; P: p= 0.009; G: p= 0.02). Three variables increase the results: being a solist (F: p= 0.046; E: p= 0.002; P: p= 0.056; G: p= 0.006), being an amateur (F: p< 0.001; E: p= 0.019; P: p< 0.001; G: p= 0.001), having a vocal complaint (F, E, P et G: p< 0.001).

This version of the V.H.I. is reliable, valid and adapted to the population of classical singers.