In this paper the ratio-type estimator &ymacr;δ = (&ymacr;/&xmacr;) &Xmacr; δ, is studied from the viewpoint of pξ-inference in order to analyze if the gain in efficiency obtained by &ymacr;δ with respect to &ymacr;R can be considered significant or not. This study involves two superpopulation models under simple random sampling design without replacement.
