CONICET | Buscador de Institutos y Recursos Humanos

A pool of neuropsychological test was selected to build a semantic memory assessment battery. The criteria used to select the test were: a) to include items of both living and non-living domains; b) verbal and pictorial stimuli; c) taxonomic and thematic conceptual relations; d) stimuli with high and low familiarity; e) tests that provide semantic cues; f) tests that were culturally build or adapted in Argentine. The battery was composed of the following tests: PAPDI (naming test), TAXON (test to assess taxonomic relations), Pyramids and Pharaohs (the Argentinean version of Pyramids and Palm tree), Sentence verification of the EMSDA (Argentinean version of the Semantic Memory Battery for Alzheimer Disease) and auditory word comprehension from the BEA (Aphasia Assessment Battery). The sample included 27 healthy participants and 31 patients (19 with focal brain damage and 12 with degenerative disease). The factorial analysis indicated that only one dimension explained 68% of the variance. Internal consistency was assessed and significant correlations were observed between all the tests (p < .001) with the exception of TAXON and PAPDI (rs = .243; p = .086) and Sentence Verification and Pyramids and Pharaohs (rs = .236; p = .096). Concurrent validity was assessed through the correlation of the total battery score with Synonym Judgment from the BEA (rs = .489; p < .001). The highest correlations were found with the verbs both with high (rs = .677; p < .001) and low imageability (rs = .681; p < .001). Criterion validity was evaluated through a comparison between patients and healthy participants (U = 88; p < .001). In conclusion, evidences of concurrent and criterion validity were obtained. Internal consistency was observed with some exceptions. Nonetheless, it was expected to find some discrepancies between tests as long as the battery assess different aspects of semantic memory.