Rejoinder on: An updated review of Goodness-of-Fit tests for regression models

Abstract

First for all, we would like to thank the discussants for reading our paper and for taking time to prepare such interesting and valuable contributions. The feeling, after revising the discussions and going back to the original version of the paper, is that there is still too much to say about Goodness-of-Fit (GoF) tests for regression models and the discussants have given a good proof of this. Although they qualify the review as thorough and nearly exhaustive, it is clear after their comments there are quite a few issues and open problems that were left untreated.We have tried to organize this rejoinder according to the different topics that have been brought up in the discussions, instead of answering each discussion separately. Bearing this in mind, some brief comments will be made on (a) tests based on the error distribution; (b) tests with dependence between error and covariate; (c) bandwidth selection and calibration; (d) non-and semiparametric hypotheses; (e) tests in survival analysis and (f) some further topics.(a) Tests based on the error distribution In Sect. 2.4, concerned with this topic, two main methodological approaches are presented. The first one, collected in the paper by Van Keilegom et al. (2008), is based on the empirical distribution of the residuals, whereas Huskova and Meintanis (2009) considered an alternative route This rejoinder refers to the comments available at