van de Vijver and Leung (1997) have discussed two additional types of equivalence, both of which can probably be considered as subtypes of metric equivalence: measurement unit equivalence and scalar or full-score equivalence. Both of these concepts are worthy of brief consideration because they are important for both theoretical and applied cross-cultural uses of psychological tests.
This level of equivalence indicates that a measure that has been adapted for use in a target culture continues to have the same units of measurement in the two culture-specific forms. That is, both forms of the measure must continue to yield assessments that follow an interval scale, and in addition, it must be the same interval scale. If a translated form of a test were studied using a sample in the target culture comparable to the original norm group and the new test form was found to have the same raw-score standard deviation as the original, this finding would be strong evidence of measurement unit equivalence. If the norms for the target population were extremely similar to that in the original population, these data would also be extremely strong substantiation of measurement unit equivalence.
Scalar ( full-score) Equivalence
Scalar equivalence assumes measurement unit equivalence and requires one additional finding: Not only must the units be equivalent, the zero-points of the scales must also be identical. Thus, the units must both fall along the same ratio scale. It is unlikely that many psychological variables will achieve this level of equivalence, although some physiological variables, such as birth weight, certainly do.
Was this article helpful?