The Vital Role of Measurement Equivalence in Family Research

The construct is the most basic aspect of a theoretical model. Whether a construct is equivalent across groups or time is a vital consideration when testing theory. However, examining construct equivalence is often not well understood either conceptually or statistically. This article provides a conceptual and statistical overview of construct equivalence as well as a rationale for its increased use. An example of testing for equivalence is also provided. Three hundred 12-year-olds rated their mother and father on a scale of parental engagement. It was examined whether this scale demonstrated equivalence for both mothers and fathers. Equivalence tests found that strong equivalence was not met, and therefore mean comparisons are not justified. Theoretical and practical implications of measurement equivalence are discussed.