One Variance

Confidence intervals - Bonett method

  

Use the Bonett method if your sample data are continuous, but do not come from a normally distributed population. These two confidence intervals provide ranges of values that are likely to contain the unknown population standard deviation and variance, respectively.

Example Output

Method

 

The chi-square method is only for the normal distribution.

The Bonett method is for any continuous distribution.

 

 

Statistics

 

Variable   N  StDev  Variance

Length    50  0.871     0.759

 

 

95% Confidence Intervals

 

                                         CI for

Variable  Method       CI for StDev     Variance

Length    Chi-Square  (0.728, 1.085)  (0.529, 1.178)

          Bonett      (0.704, 1.121)  (0.496, 1.257)

Interpretation

In the lumber yard example, if the data are nonnormal, then the standard deviation of the length of all beams from the lumber yard is likely to be between 0.704 cm and 1.121 cm. The variance of that length is likely to be between 0.496 cmimage\squared.gif and 1.257 cmimage\squared.gif.