TABLE OF CONTENTS Since the seventies of the twentieth century, the allure of nonparametric models has become almost irresistible in statistical methodology due a number of factors. Among those are the rise of the discipline of Data Analysis and of random estimation procedures spearheaded by J.Tukey, P.J.Huber and F.Hampel and unprecedented advances in computer technology. Psychologically these models are quite appealing since they free us from the constraints of distributional "straight jacket". However, the lesser efficiency and ambiguity of the results sometimes turns out to be quite heavy.
This chapter deals with the nonparametric stress-strength model where the distributions of X and Y are unknown. The version of the problem is quite important not only because it is the only set-up that can be used in a number of applications but also since it preceded historically the parametric formulation of the problem.
In this chapter we shall essentially follow the same well trotted route as in earlier three chapters of the book. We start with construction of the point estimator 
of R= P(X. In Section 5.3 we shall provide confidence intervals for R based on . Section 5.4 is devoted to nonparametric Bayesian approach to the problem. Finally, Section 5.5 describes a probabilistic design approach to the problem.
Chronologically, the stress-strength model started in a nonparametric setup in the simple...
