Abstract:Parametric hypothesis testing associated with two independent samples arises frequently in several applications in biology, medical sciences, epidemiology, reliability and many more. In this paper, we propose robust Wald-type tests for testing such two sample problems using the minimum density power divergence estimators of the underlying parameters. In particular, we consider the simple two-sample hypothesis concerning the full parametric homogeneity of the samples as well as the general two-sample (composite) hypotheses involving nuisance parameters also. The asymptotic and theoretical robustness properties of the proposed Wald-type tests have been developed for both the simple and general composite hypotheses. Some particular cases of testing against one-sided alternatives are discussed with specific attention to testing the effectiveness of a treatment in clinical trials. Performances of the proposed tests have also been illustrated numerically through appropriate real data examples.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to improve the robustness of the test method when conducting parametric hypothesis tests between two independent samples. Specifically, the paper proposes new two - sample Wald - type test methods based on the Minimum Density Power Divergence Estimator (MDPDE). These methods aim to solve the following types of problems:
1. **Complete homogeneity test**: Test whether two samples are from the same population, that is, test whether the parameters of the two samples are exactly the same. For example, for normally distributed samples, test whether the means and variances of the two groups of samples are equal:
\[
H_0: (\mu_1, \sigma_1) = (\mu_1, \sigma_2) \quad \text{vs} \quad H_1: (\mu_1, \sigma_1) \neq (\mu_2, \sigma_2)
\]
2. **Partial homogeneity test**: In the presence of nuisance parameters, test whether certain specific parameters are equal. For example, when the variances are unknown but the same, test whether the means of the two groups of samples are equal:
\[
H_0: \mu_1 = \mu_2 \quad \text{vs} \quad H_1: \mu_1 \neq \mu_2
\]
3. **One - sided alternative hypothesis test**: Pay special attention to the test of treatment effects in clinical trials, for example, test whether a new drug is more effective than existing drugs.
The paper improves the robustness of the test method in the presence of outliers or data contamination by introducing MDPDE to replace the traditional maximum likelihood estimation (MLE). MDPDE balances the efficiency and robustness of the estimate by adjusting a non - negative tuning parameter \(\beta\). When \(\beta = 0\), MDPDE degenerates into MLE, and when \(\beta>0\), MDPDE has better robustness.
The paper also studies in detail the asymptotic distributions and theoretical robustness of these Wald - type test methods, and verifies the effectiveness of these methods through numerical simulations and real - data examples.