Abstract
It is argued in response to Bornstein et al. (1983a) that most preferreed outcomes are not restricted to the maximal values of strategies, that the revised matrices are conpromise between maximal discrimination and exact fairness, that strategies can and should be measured as variables, and that their further criticisms of the Taifel matrices again fail to withstand careful examination. Their new data are irrelevant to the measurement issue. It is concluded that the Tajfel matrices are superior to the revised matrices if one wishes to measure strategies as unconfounded variables and without defining subjects' most preferred outcomes a priori.

This publication has 5 references indexed in Scilit: