Quantile regression coefficient modeling for counts to evaluate the productivity of university students

The extension of quantile regression to count data raises several issues. In this research we compare a solution which exploits jittering to obtain a continuous working variable to a more recent approach, in which the coefficients of quantile regression are modeled by parametric functions. Both methods are applied for evaluating the effect of remote teaching on university students’ productivity. In this context, the latter approach is found to be advantageous.