STEM professional development programs for in-service teachers: a meta analysis
Overview: This project evaluates the effectiveness of STEM professional development programs for in-service teachers. It focuses on measuring improvements in teachers' content knowledge, beliefs, and pedagogical skills. By synthesizing findings from multiple studies, the project aims to identify which program features are most beneficial for enhancing STEM teaching practices.
My contributions:
Performed a meta-analysis using Python to aggregate effect sizes (using metrics such as Cohen's d and Hedges' g) across diverse studies. Analyzed heterogeneity among study outcomes and conducted moderator analyses to pinpoint key factors influencing program effectiveness.
Publication:
STEM Struggle Detection Initiative (undergoing)
Overview: Using big data analytics to identify students who are struggling in STEM-related majors by analyzing historical data.
Techniques: Applying k-means clustering and regression methods to segment student populations and predict those at risk.
Expected Outcomes: Early identification of students in need will allow for timely intervention, targeted support, and ultimately higher retention and success rates in STEM programs.
My Contribution: I am leading the design and implementation of the analytical framework, integrating diverse data sources to generate actionable insights that inform intervention strategies.
Publication:
AI-Driven Early At-Risk Student Identification (undergoing)
Overview: Developing an AI-powered tool that analyzes historical data alongside current behavioral, background, and performance metrics to detect at-risk students early.
Techniques: Combining machine learning models and data integration strategies to create predictive models that flag students in need of additional support.
Expected Outcomes: The tool will provide early warnings, enabling institutions to offer proactive support, with a particular focus on advancing educational equity for underrepresented groups.
My Contribution: I am spearheading this project by conceptualizing the model architecture and overseeing the integration of multi-source data, ensuring the tool delivers interpretable and actionable insights to improve student outcomes.