Escala de avaliação da metacognição infantil: elaboração dos itens e análise dos parâmetros psicométricos

Abstract

Metacognition represents thoughts and knowledge that individuals have about their own thoughts and cognitive processes. Two basic elements of metacognitive functioning are metacognitive knowledge, characterized by awareness of the processes and skills required to perform a certain task, and control or cognitive self-monitoring, which includes judgments about the effectiveness of the strategies initially proposed, and if necessary, the establishment of new ones. Given the lack of brazilian instruments that measure this ability, this study focused on the development of items and the analysis of evidence with respect to the validity and reliability of a scale for assessing metacognition in children between nine and 12 years of age. The first phase involved the construction of Likert-style Metacognition Scale, using a six points rating scale for the respondent to indicate the extent to which he would using each strategy. During the second phase, three experts evaluated the instructions and the 70 items to verify the content validity of the new scale. These experts were university professors with doctoral degrees, and two were specialists in metacognition. The third judge, given her expertise in elaboration of scales, assessed the adequacy of the instructions for the task. The reliability analysis performed on the responses of the experts presented a 70.3% of agreement. The specialist in elaborating scales suggested some modifications in the instructions, which are incorporated into the scale. Two samples of each which led to further changes in the instructions and reduction in the number of items from 70 to 67. The factor analysis performed on large sample of 196 participants revealed the presence of a single factor denominated Metacognition , responsible for 16.84% of the variance in the scale. After exclusion of items that had factor loadings below 0.30 of items that indicated misunderstandings by the participants and similar items, the new version of the EMETA was composed of 40 items with an internal consistency of 0.90. There were no significant differences related to performance variables such as gender, school type and age of the participants. As such, the analysis of validity and reliability of the items indicated adequacy of the concept initially proposed and good internal consistency of the final version of the scale.