| Abstract: | A common procedure that is frequently used in perceptual studies interested in group differences (e.g., young vs. aged) involves initially equating all subjects in a pretreatment condition by differentially weighting the subjects on a particular stimulus dimension. The experimental treatment of interest is then introduced to these equal-baseline groups. The present article questions the general validity of this procedure by focusing on the critical assumption that the pretreatment weighting and the experimental treatment are independent. A research example is included demonstrating that differentially weighting individuals in order to produce equal-baseline performance may, in fact, provide the basis for later performance differences under the experimental treatment. |
