Key points are not available for this paper at this time.
This paper presents a bivariate response heteroskedastic regression model with flexible distributions based on the Clayton and Frank copula functions. The aim of the regression is to explain the association between the response variables, and also capture the influence of this association from two systematic components. The parameters are estimated by maximum likelihood, and several simulations are carried out to study the accuracy of the estimators. The proposed method is illustrated using data from an experiment to study the effects of different chemical seed treatments on plant nodulation. The results show that the chemical treatments do not harm the nodulation of the plants and can be applied up to 30 days before sowing.
Vasconcelos et al. (Tue,) studied this question.