The aim of this survey is to perform a validity assessment of the "Children Depression Evaluation Scale" from Amaral and Barbosa (1990). Through the research, the initial scale has been modified, resulting in four versions. It has been carried out a semantic analysis of the items with two samples of children (N= 12 and N= 4). Subsequently, four Clinical Psychologists and three Psychiatrists analyzed the pertinence of the items to the construct. Finally, the scale was applied to the 362 students from public and private schools, for the study of the precision and the validity of the construct. It was used the exploratory factorial analysis; the discriminative power index of the items, standard-deviation and Pearson's item-total corrected correlation; and the Cronbach's Alfa Coefficient calculation (a). In the semantic comprehension of the items, one item has been excluded and three items reformulated. There has been no disregard items in the judges analysis. In the exploratory factorial analysis it has been observed that the first four dimensions explained 33,36% of the total variability of the data, with self values over 1.4. Four items were discarded due to presenting factorial-load under 0.4. In the discriminative power analysis of the items, six items were eliminated with total-item correlation close to zero. The final scale comprised 24 items distributed according to four factors: disforical affection response (a= 0,70); emotional-affective response with anxiety contort (a= 0,65); autonomous-physiological response (a= 0,56), and cognitive response with self-esteem content (a=0,50).
Child depression; psychological assessment; instruments; mesurement
