Title: Inference in nonparametric current status models with covariates

Abstract: In interval censored models with current status observations, the variables are indicators of the presence of individuals on observation intervals and covariates. When several individuals share the same observation interval, a simple procedure provides new estimators for the distribution of the observation times and their intensity, in a closed form. They are $n^{1/2}$-consistent for piece-wise constant covariates. Estimators of the sample-sizes are deduced and asymptotic $\chi^2$ tests for independence of the observations on consecutive intervals and for independence between consecutive classes for the observed individuals are proposed.