Evaluating the Measurement Validity of Lifestyle Typologies with Qualitative Measures and Multiplicative Factoring

The validity of two lifestyle typologies, SRI's Values and Lifestyles system and the Lastovicka et al. Drinking-Driving typology, are examined empirically. Each typology is measured by three methods. One measurement method, characterized as "quantitative," consists of statistical modeling of responses to Likert-type items collected in structured interviews. The two other measurement methods, characterized as "qualitative," are judgmental codings of data collected from a schedule of open-ended and projective tasks conducted in face-to-face, in-home interviews. Two well-known methods, Campbell and Fiske's multimethod-multitrait approach and the analysis of covariance structures (LISREL) approach, are used to examine the convergent and discriminant validity of the lifestyle typologies. In addition, Browne's class of direct product factor models is introduced and applied. In contrast to the well-known LISREL approach, direct product models assume the effects of method factors on trait factors are multiplicative instead of additive.