This book provides alternatives to current methods - alternatives that stay much closer to the original data. Thus, the book's approach can be looked on as a turn to simple empiricism in psychological measurement. Since the empiricism is based on ordinal rather than interval scale concepts, so it is more realistic than current approaches or older ones, such as classical test theory. Behavioral data is inherently noisy, so the authors emphasize providing summary scores and information on the relative amount of noise there is in the data, rather than imposing a parametric model on it.