If a predictive relationship was studied, was the predictor variable the only criterion used to select participants in the study? Would combining the predictor variable with other screening criteria improve its predictive validity? (A predictive validity coefficient of about 0.8 is needed for an accurate prediction.)