Tests regarding parameters of several independent gamma populations

Gamma distribution has been used with considerable success in reliability studies and life testing experiments. This distribution assumes various forms as its shape parameter varies which makes it suitable for analyzing a variety of lifetime data. In such studies, often interest lies in comparing a lifetime distribution over multiple groups. In this paper, we develop asymptotic tests for comparing shape parameters of k independent gamma distributions. We also develop similar tests to compare both, the shape and scale parameters, simultaneously. The tests are based on generalized minimum chi-square procedure. This procedure has been known to produce estimators which are asymptotically e±cient and the tests based on such estimators are known to have high asymptotic power.