Least squares tests for discrete parameter stochastic processes
Paul I. Nelson · Communication in Statistics- Theory and Methods · 1979
A rest tor a subset oi parameters based on least squares estimates from dependent data is developed The asymptotic distribution of a test statistic defined Lo be the difference between a sum of squares minimized under the full and reduced models is shown to be that of a linear combination of noncentraL chi-square variates. Application is made to a suocnuicai branching process with immigration.