Using Correspondence Analysis And its Distance To Evaluate The Components of A Naming Test For Studying Aphasia

Exploratory Multidimensional Data Analyses were used to manage two components of a naming test for studying lexical access in aphasic patients, i.e., the naming agreement of images and age of acquisition of the names themselves from an original international test. In order to be reliable the images should be easily and unequivocally named by any subject using the same word. Theoretical assumptions about word le arning states that words acquired later tend to be the first to be lost due to brain damage in aphasia. Thus, these two variables are important predictors of the patient’s word retrieval. We first selected the images according to normal judges recognition agreement; then, to range them based on their primitiveness, these images were submitted to two sets of judges, that had to answer according to two different scales. Data were analyzed with several exploratory multidimensional techniques, including Simple and Multiple Correspondence Analyses, Principal Component and Multiple Factor Analyses. A comparison suggested that no mayor differences existed due to the two scales’ differences.