Should ratings of the importance of personal values be centered?

Research on personal values is based on persons’ ratings of the importance of values. Typically, the means of these ratings are discarded as response style artifacts through centering the data, person by person. We show that centering leads to more circular value configurations with lower Stress in MDS than using raw data. For unfolding models, we show that using raw data avoids some special issues in unfolding; the model space requires one additional dimension; after appropriate rotations, the value circle emerges in a plane; the persons’ scattering about this plane corresponds to their mean ratings. The mean ratings correspond to the first principal component of the value items. It is demonstrated that mean ratings can also be substantively meaningful.