Testing marginal homogeneity of a continuous bivariate distribution with possibly incomplete paired data (Q2174524)

Let \[ \left[\frac{X_{1,1}}{X_{2,1}}\right],\ldots,\left[\frac{X_{1,n}}{X_{2,n}} \right] \] be a sample of size \(n \in \mathbb{N}\) of independent and identically distributed bivariate random vectors with unknown bivariate distribution function \(F\). Denote by \(F_{1}\) and \(F_{2}\) the first and second marginal distribution of \(F\), respectively. The author considers the full nonparametric testing problem of marginal homogeneity \[ \mathcal{H}: F_{1} = F_{2}, \quad \mathcal{K}: F_{1} \ne F_{2} \] with possibly missing data. He assumes that the missing components are missing completely at random (MCAR). From the author's abstract: ``Applying the well-known two-sample Crámer-von-Mises distance to the remaining data, we determine the limiting null distribution of our test statistic in this situation. It is seen that a new resampling approach is appropriate for the approximation of the unknown null distribution.We prove that the resulting test asymptotically reaches the significance level and is consistent. Properties of the test under local alternatives are pointed out as well. Simulations investigate the quality of the approximation and the power of the new approach in the finite sample case. As an illustration, we apply the test to real data sets.''