Development of the Texas Spanish Naming Test: A Test for Spanish-Speakers

The elderly Hispanic population is growing at a rapid pace, although very few neuropsychological measures are developed in Spanish, as most tests are translated versions of their English-language originals. The construct validity of these instruments with Spanish-speakers is virtually unexamined and bias may result due to cultural differences between the population the original tests were intended for and the Spanish-speaking populations they are used with. The current investigation used culturally salient words to develop the Texas Naming Test (TNT), a confrontation naming test for Spanish-speakers. Eighty-five (55 nondemented and 30 demented) Spanish-speaking primary care clinic patients were administered this test to determine its psychometric qualities. The TNT demonstrated very good internal consistency (alpha = 0.9) and good convergent validity, as it correlated highly with translated Spanish-naming tests commonly used in clinical practice (r > 0.80). Multivariate analysis of covariance and logistic regression demonstrated that performance on the TNT was not significantly influenced by acculturation, though the test did relate to both education (r = 0.48, p< .001) and acculturation (r = 0.41, p < .001). As predicted, the TNT effectively detected differences between demented and nondemented individuals, and demonstrated a high level of sensitivity (100%) for dementia when using an optimal cut score of < 23. Furthermore, ROC curve analysis demonstrated the overall discriminant utility of the TNT was comparable to a literal Spanish translation of the Boston Naming Test (MBNT-S), but better than a translated short form of the BNT (15-SNT). Data from this investigation suggest the TNT may be clinically useful where dementia among Spanish-speakers is suspected. Further exploration is needed to determine the extent to which culturally salient words contribute to greater sensitivity.