Weak convergence for empirical processes of associated sequences

We establish a weak convergence theorem for empirical processes of stationa ry and associated random variables having the uniform marginal distribution . To carry out the proof, we develop a tightness criterion for the empirica l process constructed from any stationary sequence fulfilling a suitable mo ment inequality. We apply the result to stationary non mixing moving averag e sequences with positive coefficients. Based on this class of linear proce sses, we compare mixing and association. (C) 2000 Editions scientifiques et medicales Elsevier SAS.