Latent mean structure analysis: A technique to compare group differences in a second-order quality of life model

Abstract

The challenge in using conceptual models to measure multidimensional outcome concepts has been the inability to statistically compare group differences. An illustration of latent mean structure analysis using a second-order factor model representing quality of life is presented. Tests for the structure of the second-order factor model representing quality of life, and for its invariance across the 2 groups were based on the analysis of latent mean structures. Group means on the first-order factors could be compared and there were significant differences. Nursing researchers are often interested in studying theoretical constructs that cannot be measured directly. This analysis demonstrated a process by which means of a quality of life conceptual model could be compared between groups.