The prevalence measures of disability (for example) by single or five-years of age show considerable fluctuations due to sampling variations. Mortality rates (for example at old ages) have also statistical fluctuations. In the same way, Sullivan index could be subjected to variations which would not be related to any fundamental evolution. To have an idea of the size of such random changes, it is accustomed to calculate the variance of the indicator.

- Estimate of variance of Sullivan index
- Example of variance calculation
- Interest of variances of Health expectancies

Fri Apr 25 22:40:35 DFT 1997