Potential Abstract: This research article examines the intersection of sociopolitical factors and statistical learning science in educational settings, with a focus on the application of directed acyclic graphs (DAGs). In the postmodern era, educational research has increasingly recognized the need to consider the broader social, cultural, and political contexts that shape learning outcomes. Simultaneously, the field of learning science has made significant advancements in statistical methods, such as DAGs, to model complex relationships among variables. This article seeks to bridge the gap between these two domains by investigating how sociopolitical factors can be incorporated into statistical learning science using DAGs.
This study adopts a mixed-methods approach, incorporating both quantitative data analysis and qualitative inquiry. The first phase involves a comprehensive literature review to identify key sociopolitical factors that influence educational outcomes, such as student socioeconomic status, racial/ethnic background, and school funding disparities. Drawing on this literature, we develop a conceptual framework that integrates sociopolitical factors into statistical learning science models.
The second phase of this research employs a large-scale dataset comprised of student achievement data, demographic information, and contextual factors from diverse educational settings. We apply DAGs to model the relationships among sociopolitical factors, instructional practices, and student learning outcomes. By leveraging DAGs, we can account for confounding variables, explore causal mechanisms, and generate more robust and nuanced understandings of the complex interplay between sociopolitical factors and learning science.
Our findings shed light on the ways in which sociopolitical factors act as mediators or moderators in the relationship between teaching practices and learning outcomes. Additionally, this research highlights potential avenues for interventions and policies aimed at reducing educational disparities related to sociopolitical factors.
Potential References:
- Causality and statistical learning
- Using directed acyclic graphs to guide analyses of neighbourhood health effects: an introduction
- A recursive method for structural learning of directed acyclic graphs
- Fairness-Aware Instrumentation of Preprocessing~ Pipelines for Machine Learning
- Machine learning algorithms for social media analysis: A survey
