Construção e evidências de validade de conteúdo da “Escala de Triagem para Identificação de Sinais de Autismo : versão para Professores (TEA-Prof)”

Abstract

Autism Spectrum Disorder (ASD) is characterized by persistent impairment in social communications and interactions as well as restricted or repetitive patterns of behavior and interests. An important way of tracing ASD in school-age children is through the account of educators who spend time daily with the children. Hence, this project has the objective of describing the initial procedures of construction and verification of validity evidence of a screening tool for the identification of ASD signs in children targeted for teachers, named “Teacher Screening Scale for the Identification of Signs of Autism (TEA-PROF)”. The TEA-PROF is a self-answer, 65-item teacher-oriented scale, specifically for educators who work with autistic kids between the ages of six and twelve. The items on the scale were developed based on the behavioral categories established in the operational definition of the construct, literature reviews and on the Diagnostic and Statistical Manual of Mental Disorders- DSM-5 (APA, 2014). Overall, theoretical and empirical procedures were analyzed for the TEA-PROF, divided into the following stages: national and international literature review, scale construction and content-based validity evidence verification. The first stage, scale construction, was initially composed by an integrative national literature revision (Scielo and the CAPES Theses and Dissertation database) and international (Lilacs, PubMed, PsycINFO) that aimed to pinpoint autism identification tools destined for teachers only. As a result, seven studies were included in the revision. This procedure exposed the scarcity of such materials and the need for the creation of one on a national level. It also enabled the gathering of theoretical contributions for the construction of TEA-PROF. After the construction, in order for the investigation of content-based validity evidence, the tool, initially elaborated with 86 items, was submitted in two stages to the evaluation of six health and education professionals. To evaluate the agreement between the examiners the Fleiss Kappa value was used, according to the Landis e Kock (1977) classification: Kappa < 0 = No Agreement; Kappa 0 – 0.19 = Slight Agreement; Kappa 0.20 – 0.39 = Fair Agreement; Kappa 0.40 – 0.59 =Moderate Agreement; Kappa 0.60 – 0.79 Substantial Agreement; Kappa 0.80 -1.00 = Almost Perfect Agreement. At the end of this stage, due to the evaluation of the judges, 16 items were either excluded or grouped into another item for denoting repetitions. The scale now contained 70 items, 27 of which were yet reformulated and resubmitted to the evaluation of the same specialist judges. On the second stage of judge evaluation, 4 items were excluded due to presenting a lower than 0.60 agreement. The items with substantial agreement were aggregated to the items of the first stage of judge evaluation, totaling in a 65-item version. These were presented to the target population (teachers) for a semantic analysis, in which teachers did not agree on the comprehension of 8 items. Therefore, the items where reformulated and resubmitted to the same teachers. However, only two teachers performed the analysis. Consequently, the Cohen’s Kappa method (for when there are only two judges) was utilized, and no disagreement between the participants was found.