Discussion of 'identification and estimation of non-linear models using two samples with nonclassical measurement errors'.

We would like to congratulate the authors for solving a complex, novel and stimulating problem of identification and estimation in a measurement error context. We also thank the Executive Editor for giving us the opportunity to discuss this work. Our main contribution to this discussion will be to try to develop the necessary insight into the methodology and its properties, with the aim of allowing other readers to better appreciate their highly original approach. We also raise a few points that the authors may wish to develop. [ABSTRACT FROM AUTHOR]