In policy evaluation, there is a concern for arriving at policies that will achieve or maximize societal or group goals. In traditional social-science research, there is a concern for establishing reality, particularly about predictive causal relations, regardless of whether the findings have any policy or practice usefulness. Thus there is generally more at stake in each of these dilemmas than there are in the counterpart situations in traditional social-science research. In the average social-science study, one normally does not care relatively so much about the nine counterpart situations as to.