Least Squares Estimation for the Inverse Power Law for Accelerated Life Tests

Procedures are given for obtaining orthogonal least squares estimators for the parameters of the inverse power law with an exponential failure distribution when the data are obtained from censored sample accelerated life tests. Prediction at use stress is also considered. Properties of the estimators are examined and the exact cumulants of the estimators are used in the Edgeworth series to obtain approximate confidence levels. These procedures yield results that are in good agreement with those based on exact and asymptotic properties of the maximum likelihood estimates.