RESUMO
The study describes the development of the AAC-Arm questionnaire and its initial psychometric and logometric testing for reliability and validity. Psychometric and logometric principles were used to develop an assessment questionnaire capable of evaluating the communication state domains important to patients with neurological disorders. The hypothesized domains were to include (1) auditory function, (2) speech function (3) cognitive functions (4) sensorimotor function, and (5) activities of daily living (ADL). An initial pool of 78 questions was pilot-tested for clarity in 10 patients; following factor analysis, the number of questions was reduced to 39-items. Then the questionnaire was subjected to reliability and validity testing. Factor analysis supported the 5 hypothesized domains. Test-retest reliability using Spearman's correlation demonstrated substantial agreement, ranging from 0.72 for the ADL domain to 0.92 for the auditory function domain. In testing for internal consistency, Cronbach's alphas ranged from 0.86 for-the ADL domain to 0.96 for the cognitive function domain. Correlation between domains gave evidence of construct validity. In comparing similar domains in the AAC questionnaire, a moderate correlation (range 0.33-0.83) for the ADL and sensorimotor function scales were found. The correlation was more positive between the other domains. Testing of reliability for the phraseological, syntactic and semantic competence indices showed good positive correlation between initial and retest scores. The questions in the AAC questionnaire have undergone rigorous psychometric and logometric testing, and the tool is an appropriate instrument for the assessment of neurological patients with communication deficit. The psycholinguistic assessment provides with the main weight of data for successful communication therapy.