THE IMPACT OF ACADEMIC PERFORMANCE ON THE FORMATION OF FRIENDSHIP NETWORKS AMONG HIGH SCHOOL STUDENTS IN THE METROPOLITAN REGION OF NATAL/RN"
academic performance; social capital; sociocentric network analysis; Social Selection; public high school.
Context and Problem: Given the educational challenges, and despite Brazil's academic performance indicators falling below international standards, the Northeast region and the state of Rio Grande do Norte frequently show indices of academic performance below national expectations. Beyond macrostructural analyses, it becomes essential to investigate the micro-social mechanisms operating in the school environment, where the social context and friendship networks act as factors in the reproduction or mitigation of inequalities. Goal and Hypothesis: From this context, the present study aims To analyze how academic performance acts in the formation of ties through generative processes (homophily and transitivity) that operate within student networks in different public high school contexts in the Metropolitan Region of Natal (RMN/RN). The study starts with the hypothesis that academic performance, besides being a resulting variable, is a strong predictor of network selection by similarity (homophily), thus perpetuating patterns of social interaction and educational inequalities. Methodology: The proposed methodology adopts a quantitative and cross-sectional approach, structured in two stages: 1) a systematic literature review (in the Web of Science, Scopus, SciELO, and PsycINFO databases) to map the associations between networks and academic performance, and 2) the analysis of secondary data collected via questionnaires applied to students ($N=177$) in three public schools in the RMN/RN in 2015, during the extension project "O habitus de estudar: construtor de uma nova realidade na educação básica da Região Metropolitana de Natal". Differential and Expected Results: The methodological differential of this study resides in the construction and analysis of the students' sociocentric networks and the use of Exponential Random Graph Models (ERGM) to model the formation of ties, controlling the interference between the processes of homophily (by academic performance) and transitivity (tendency to close triangles). It is expected to confirm that academic performance is a strong determinant of tie formation, especially by similarity—an assumption made more crucial given the high aspirational homogeneity (100% in two schools) and the heterogeneity of parental cultural capital observed in the sample, reinforcing the need to investigate social selection at the micro-level.