PAPDIC: a new color naming test

LINARES, N.; VIVAS, L.; MANOILOFF, L.; FERNÁDEZ ZAIONZ, A.

Chicago

Conferencia; Alzheimer´s Association International Conference; 2018

Alzheimer´s Association

Background: The Argentinean Psycholing?uistic Image NamingTest (PAPDI) is a test that was developed initially in black and white and more recently in color. An important characteristicof this test is that pictures and semantic cues were selectedconsidering psycholinguistic variables from Argentine normativedatabases. The test is comprised of 30 color images extractedfrom Cycowicz et al. (1998) set comprising living andnon-living things. Psychometric properties of the new colorversion (PAPDIC) were analyzed. Methods: Clinical sample:28 patients (18 left brain focal damaged and 10 with neurodegenerativedisease). Normative sample paired in age and educationallevel: 26 subjects without neurological impairment.Subjects were administered a neuropsychological assessmentbattery including language and semantic memory tests. Psychometricanalysis performed: internal consistency was assessedthrough Kuder-Richardson (KR20) as items were dichotomic;concurrent validity with Spearmans Rho with the brief versionof the Boston Naming Test; criteria validity through groupcomparison (clinical vs control) with Mann-Whitney U test;sensitivity and specificity to detect anomia through ROCcurves; and the discriminating power of items through point-biserialcorrelations with the total test score. Results:The PAPDICshowed a highly acceptable internal consistency value (.927) anda significant correlation with the Boston Naming Test(Spearmans Rho = .789; p < .001). Significant group differenceswere observed (z = -5.242; p < .001). ROC curve analysisshowed a Sensitivity of 90% and a Specificity of 100% with anArea Under de Curve of .984 (IC 95% .949-1). The suggestedcutting point according to Youden Index was 26. Regarding thediscriminating power of individual items, it was observed thatthe following pictures showed the highest point-biserial correlations:envelope (rpb = .844), brooch (rpb = .798), and fishbowl(rpb = .773). Conclusions: The PAPDIC showed internal consistencybetween items. Evidence of concurrent and criteria validitywere obtained. Specificity and Sensibility values werehighly acceptable and showed better discriminating valuethan the B&W version (AUC ¼=.817). A cutting point to detectanomia was established. Items with better discriminating capacitywere identified. Results are promissory and suggestthat the PAPDIC could be a useful tool to assess naming difficulties.