Research Article
An Empirical Analysis of the List of Values (LOV) Scale Using the Ipsative Factor Model
Published: January 1995 · Vol. 24, No. 1 · pp. 317-349
Full Text
Abstract
Ordinal or ranking scale data frequently used in marketing research possess ipsative properties, making conventional factor analysis inappropriate. In this regard, this methodological study introduces the ipsative common factor model (Jackson & Alwin 1980; Alwin & Jackson 1982), which is relatively unfamiliar in the marketing field, and examines its broad applicability using LOV (List of Values) ranking and rating scale data collected from Korean, American, and Japanese university student samples. The empirical analysis using the ipsative single-factor model demonstrated that for the Korean sample, the data fit of the proposed model was confirmed, and parameter estimates were relatively consistent with prior research results using rating scales. However, the American and Japanese samples yielded comparatively unsatisfactory results. Furthermore, the transformation matrix (A = I - p⁻¹ll') assumed by Jackson and Alwin for converting between ipsative and non-ipsative scales was found to be inadequate, and the need for an alternative transformation matrix was noted. Finally, the limitations of this study and implications for future research were presented.
