¿Pirámides y Palmeras o Pirámides y Faraones?: Adaptación y validación de un test de asociación semántica al español rioplatense

Abstract

La memoria semántica es definida como el sistema que permite almacenar el significado de las palabras, objetos, conceptos y el significado del mundo en general. El test más utilizado para evaluar los déficit semánticos adquiridos es el Test de Pirámides y Palmeras (Howard & Pat - ter son, 1992). Es una prueba de asociación semántica que se administra desde diferentes modalidades (pictórica y verbal) y se encuentra muy condicionada por el medio socio cultural. Los objetivos de este trabajo son: presentar la adaptación y validación del Test de Pirámides y Palmeras al español rioplatense, comparar una versión original del test con una nueva versión abreviada y modificada, denominada Test de Pirámides y Faraones, comparar el rendimiento entre sujetos normales y pacientes y obtener puntajes de corte para el diagnóstico de dificultades semánticas en ambas versiones. El test se administró en forma computarizada a 50 voluntarios (40 controles y 10 pacientes con demencia semántica). De las 66 tríadas adaptadas y construidas se seleccionaron las 20 que mejor discriminaron entre pacientes y controles. La fiabilidad del test original fue de alpha igual a .857 para la modalidad pictórica y alpha igual a .910 para la modalidad verbal. La del nuevo test fue de alpha igual a .917 para la modalidad pictórica y alpha igual a .918 para la modalidad verbal. Se obtuvo así la adaptación y validación del Test de Pirámides y Palmeras, co mo así también una versión abreviada, el Test de Pirámides y Faraones, que evalúa adecuadamente la memoria semántica en nuestro medio sociolingüístico y se adecua a los requerimientos clínicos actuales.

Semantic memory is a long term memory system proposed by Tulving (2000) that stores objects, words, and general world knowledge’s meanings without connection with any particular time or place. Conceptual knowledge is mostly shared across individuals in a given culture, although its precise scope depends on the individual’s experience (Hodges & Patterson, 1997; Patterson & Hodges, 1995). Semantic memory may be impaired in many neurological disorders. This disruption may be attributed to pathology in the infer-lateral temporal lobes. Patients with semantic dementia have difficulties with objects and words meanings (Budson & Price, 2005). Pyramids and Palm Trees Test is one of the most used measures to assess acquired semantic impairments (Howard & Patterson, 1992). It’s a semantic association test and has six different administration modalities: pictorial, verbal, and combined. This test contains 52 triads. The English normative data from the original Pyramids and Palm Trees Test Manual (Howard & Patterson, 1992) was only obtained in 13 young adults, and no participant made more than three errors. This is a socio-cultural influenced test. The aims of this article are to present the Pyramids and Palm Trees Test adaptation and validation to our language (Spanish) and cultural context, to compare the 52 triads from the original version with a new and shorter 20 triads version, to assess differences in performance between controls and patients in both tests, and to get cutoff scores on both versions. A computerized version of the original test (52 triads) plus 14 new triads (66 triads in total) were administered to 50 volunteers (40 controls and 10 semantic dementia patients). Presentation pro - gram was used to present the stimulus. Non frequent cultural associations were omitted: (a) windmill, tulip-daffodil, (b) carrot, lamb-donkey, (c) acorns, donkey-pig, and (d) Eskimo-rowing, boat-kayak. Also, others triads were slightly modified: (a) caterpillar, butterfly-dragonfly by caterpillar, butterfly-ant, (b) Eskimo, igloo-house by Indian, carp-house, (c) crook, sheep-mice by dog, rabbit-mice, (d) padlock, bicycle-car by pump, bicycle- car, (e) blackboard, table-desk by blackboard, pen-chalk, (f) eggs, hen-swan by flock, hen-duck, and (g) soldier, church-castle by knight, church-castle. Triads with composed words in Spanish were changed: (a) safety pin (alfiler de gancho), girlbaby by pacifier, girl-baby, (b) safe (caja fuerte), necklace-tie by jacket (chaleco), necklace-tie, and (c) bath, owl-woodpecker (pájaro carpintero) by bath, owl-canary. Of the 66 adapted triads, the 20 that allowed better discrimination between pa - tients and controls were selected. The new and shorter version is called Pyramids and Pharaohs, because the Pyramids and Palm Trees Test triad had low specificity and moderate sensitivity in our sample and wasn´t selected. In the adapted Pyramids and Palm Trees Test the reliability index of the pictorial version was moderately high (α = .857), and high for the verbal modality (α = .910). In the Pyramids and Pharaohs Test the reliability index was high for both versions (pictorial: α = .917; verbal: α = .918). The cutoff score for the original version was 44 for the pictorial modality and 43 for the verbal modality. In the Pyramids and Pharaohs Test the cutoff score was 17 for the pictorial modality and 18 for the verbal one. Regarding the specificity, the adapted Pyramids and Palm Trees Test was high (98.8%) same as the new shorter test. In relation to the sensitivity, the original test was moderate (70%), lower than the Pyramids and Pharaohs Test (85%). Results indicate that the Pyramids and Palm Trees Test can be considered an appropriate adaptation to our social culture. Moreover a new test was designed, Pyramids and Pharaohs, with only 20 triads, adequate for semantics acquired impairments assessment, useful for the research on cognitive processes and current clinical re - quirements.