Measurement Invariance

Broadly speaking, measurement invariance refers to the extent to which the psychometric properties of a measure remain constant across multiple conditions (e.g., groups of respondents, time periods, [Page 919]medium of measurement administration). The opposite of measurement invariance is known as differential functioning (DF) and is representative of a lack of invariance across samples. Methods of detecting a lack of invariance have received increased interest over the past three decades, because the presence of invariance is considered an essential prerequisite to making meaningful cross-group comparisons. If measures cannot be shown to be invariant, some comparisons of scores across conditions can be uninterpretable. Furthermore, the presence of DF indicates that an item or scale may not accurately represent the latent trait of interest for at least one ...