RESUMO
The application of the Boston Naming Test (BNT) is time-consuming and shortened versions need to be developed for screening purposes. The aims of this study were to develop four equivalent 15-item forms of a Spanish adaptation of the BNT, to test the equivalence of the new versions in a clinical sample, and to provide normative data. The normative sample consisted of 340 subjects. The clinical sample included 172 patients (76 Mild Cognitive Impairment and 96 Alzheimer's disease). An empirical procedure was used to develop the shortened versions. All new versions demonstrated satisfactory internal consistency. Pearson's coefficient analysis showed strong relationships among the four short-form versions as well as between each of them and the 60-item test. The inferential confidence interval method demonstrated the equivalence between the four shortened versions. Age and education affected the score of all short-form versions, but sex was found to be unrelated to the performance. Normative data were calculated for midpoint age groups. This paper proposes four 15-item equivalent versions that could be useful and time-saving tools for screening purposes.