Some characterizations of geometric type distributions based on order statistics (Q1095520): Difference between revisions

Let \(X_{1:n}\leq X_{2:n}\leq...\leq X_{n:n}\) denote the order statistics, and \(S_{m,r}=X_{r:n}-X_{m:n}\) be a spacing of order r- m, \(1\leq m<r\leq n\), obtained from a random sample of size n from a discrete distribution. We obtain several characterizations of distributions of geometric type in the class of discrete distributions based on the properties of \(S_{m,r}\) and \(X_{m:n}.\) Specifically, these are based on (i) independence of \(X_{1:n}\) and \(S_{1,r}\), for some \(r\geq 1\), (ii) conditional independence of \(X_{m:n}\) and \(S_{m,r}\) conditioned on the event \(\{S_{m-1,m}>0\}\), for some \(m\geq 1\), and (iii) equality of the distribution of \(1+X_{r- 1:n-1}\) and the conditional distribution of \(S_{1,r}\) given \(X_{1:n}\) and the event \(\{S_{1,2}>0\}\).