Factorial Structure of the Family Values Scale From a Multilevel-Multicultural Perspective

Abstract

In cross-cultural research, there is a tendency for researchers to draw inferences at the country level based on individual-level data. Such action implicitly and often mistakenly assumes that both the measuring instrument and its underlying construct(s) are operating equivalently across both levels. Based on responses from 5,482 college students sampled from 27 countries, we took a structural equation modeling approach to addressing this issue of level equivalence. Purposes of the study were: (a) to validate the hypothesized two-factor structure of the Family Values Scale (FV Scale; Georgas, 1999) within a multilevel framework that took individual- and country-level information into account; (b) to test equivalence of the FV Scale across individual and country levels; and (c) to evaluate relations between the FV Scale and three possibly important covariates—gender at the individual level, and affluence and religion at the country level. Implications of findings and importance of multilevel equivalence in cross-cultural research are discussed.

Keywords:

• cross-cultural research
• measurement/structural equivalence
• multilevel modeling
• structural equation modeling

Notes

Although Jöreskog (1971) initially recommended that all tests of equivalence begin with a global test of equivalent covariance structures across groups, this initial test has since been disputed as it often leads to contradictory findings (for details, see Byrne, 2011).

Although this definition applies also to a full structural equation model (e.g., path-analytic model), given our focus on testing for equivalence of a measuring instrument across level, we limit our explanatory statements to the CFA model.

For categorical data, the intercepts represent the thresholds.

There is some confusion regarding both the term “metric invariance” and the assignment of Meredith's (1993) categorical labels within the cross-cultural literature where configural invariance is commonly referred to as “weak invariance,” factor invariance as “metric invariance,” and intercept invariance as “scalar invariance” (see, e.g., Davidov, 2008).

For reasons of technical complexities, the data for South Africa, Botswana, and Mongolia could not be used and, thus, these countries were eliminated from all analyses.

These analyses were based on the EQS 61 program (Bentler, 2005).

For purposes of data collection, specific religions were reported as percentages for each country. Specific to this multilevel study, the category of “Christian” included the religions of Christian Orthodox, Roman Catholics, and Protestants; the category of “Muslim” included the Islamic subgroups of Sunnis and Shias.

Values greater than 5.00 are indicative of multivariate non-normality (Bentler, 2005).

In addition to overall assumption of multivariate normality, the MUML estimator cannot be used in testing (a) cross-level equivalence, as it is not possible to calculate a χ²-difference value when country clusters are of unequal size, and (b) validity of the covariates due to an assumption of no missing data, which was evidenced for the covariate of gender.

This difference value was corrected to take into account the scaling of the MLR estimator.