Adaptação para pré-escolares de subtestes da bateria multidimensional de inteligência infantil

Abstract

The adaptation or elaboration of tests that can validly and accurately evaluate the intellectual abilities of Brazilian preschool children is relevant, considering, among other factors, the lack of instruments. The Cattell-Horn-Carroll Model (CHC) is regarded as the one that most tends to give rise to techniques of intellectual evaluation and consists of a multidimensional vision that conceives the intelligence through capacities or general factors. The Multidimensional Battery of Children Intelligence (BMI), based on such a model, is intended for the evaluation of children between 7 and 12 years. The general objective of the present study was to adapt for preschoolers and to analyze the evidence of validity and accuracy of the sub-tests General Vocabulary, Memory for Names, Speed of Perception and Response Speed of BMI. The first stage of the study consisted of the adaptation of the General Vocabulary, to be used in children between 4 years and 6 years and 11 months. Participants included 10 mothers and 10 female teachers who answered a questionnaire and the original BMI items. 227 different words of children's knowledge were raised and 86 which they were not yet able to understand, according to parents and teachers. Through the relative frequency of the words, the selection was made to compose the new items of the subtest. Thus, from the study with parents and teachers, the 1st version of the General Vocabulary subtest for preschoolers was prepared, consisting of the 35 original BMI and 18 new items. The second step consisted in sending the four subtests to five expert judges to investigate the evidences of content validity through the analysis of agreement between them. The first version of the General Vocabulary sub-test for preschoolers was sent to the judges, as well as three other original BMI subtests: Memory for Names, Speed of Perception, and Speed of Response. After the judges' analysis, the General Vocabulary subtest had 6 excluded items and 5 reformulated items. The Memory for Names only had change in one item and the other subtests did not change in their items, only in the instructions. The third step consisted of a pilot study with 10 children, so that changes in instructions and two subtests were found. Eight items of the General Vocabulary were excluded, giving rise to the 3rd version of the General Vocabulary with a total of 38 items; In Velocity of Perception the numbers were replaced by figures, giving rise to the 1st version of this subtest for preschoolers. Finally, the fourth step was the application in a sample of 43 preschoolers for the analysis of validity evidences based on relations with external variables and precision. All subtests had statistically significant results in relation to the "age" variable. As for accuracy, the 3rd version of the General Vocabulary had alpha of 0.81 and the 1st version in the Memory for Names of 0.42. In the Rapid Response and in the 1st version of the Processing Speed it was not possible to calculate the accuracy. In this way, we conclude that the adaptation to preschoolers of BMI subtests presented evidence of validity based on the variable "age" and the 3rd version of the General Vocabulary was accurate. It is necessary to carry out studies with a larger sample, in order to continue the validity and precision surveys