Exploring the normative equivalence of a measure focuses on determining if there is similarity between performance of a norm group on a particular test, and performance of a culturally distinct target group on a translated version of the same test. If the normative distributions of the two culturally different groups are not significantly different, the original norms may prove useful, with caution implied (Cuellar, 1998). Further, does the translated test discriminate to an acceptable level of confidence between individuals displaying a particular characteristic or functional ability, and those who do not? Statistically, this ability to differentiate is referred to as discriminant validity. However, more importantly, is this ability to statistically discriminate at a level of specificity and sensitivity that is clinically useful? One may find nonsignificant differences between groups regarding normative information, with several possible clinical implications: (a) the original normative data may be used with confidence; (b) the original normative data may be used, but in conjunction with a moderator variable, such as an acculturation measure, which allows development of a formula for adjustment of scores to make them meaningful (Cuellar, in press). However, in both instances, it is important to determine the sensitivity, specificity, and accuracy of the normative or adjusted normative data used. Adjustment in cutoff scores to maximize correctly classified nonimpaired and impaired participants in the target group will prove essential in making the translated instrument clinically useful in instances where there exist dichotomous clinical categories. Berry (1979,1980) recommended the use of "ethunits" which would assist adjustment of scores based on dimensions and characteristics that make cultural groups unique. Along these lines, Helms (1992) supported the use of acculturation level as an important variable to include in research on tests of cognitive ability.
If significant differences are discovered between the normative group and the culturally distinct group's performance on the translated test that cannot be useful with adjustments to scores, the original norms may not be appropriate, requiring development of new norms specific to the targeted cultural population and translated instrument. Even so, it will be important to develop cutoff scores that maximize sensitivity and specificity of the translated test to enhance its clinical utility or clinical equivalence.
Was this article helpful?