Insulating fluid. The time to breakdown data in Table 6.1 of Chapter 6 yield the seven sets of linear estimates and variance and covariance factors below for parameters of an extreme value distribution for the In data. According to theory for such data, time to breakdown has an exponential distribution at any test voltage (kV). Assess this with the following, assuming that the seven distributions are Weibull with a common shape parameter.
(a) Calculate the variance factor B for the pooled linear estimate δ* for the extreme value scale parameter.
(b) Calculate the pooled linear estimate δ and the corresponding estimate of the common Weibull shape parameter ß
(c) Calculate two-sided 95% confidence limits for δ and ß, using the normal approximation..
(d) Do (c), using the Wilson-Hilferty chi-square approximation. Compare with (c).
(e) Are the data (the pooled estimate) consistent with ß= 1 (exponential distribution)? If test conditions at a test voltage are not consistent, the p estimate tends to be lower. Are the data consistent with this possibility?