91Ö±²¥

This is a graduate-level advanced statistics course designed for students in education and the social sciences whose research involves analyses of multi-level and/or longitudinal data. Examples of multi-level data include students nested within classrooms and schools, teachers nested within schools and school districts, children nested within families and neighbourhoods, and employees nested within organizations. Examples of longitudinal data include repeated measures of child development, students' academic growth, teacher improvement, and organizational change. Multi-level modeling, also called ''hierarchical linear modeling (HLM)'', resolves the dilemma of ''units of analysis''. More importantly, it enables researchers to partition variance-covariance components with unbalanced data and to model cross-level effects with improved estimation of precision. This course will cover basic two-level and three-level models, growth curve models, and multi-level experimental and quasi-experimental designs. The objective is to equip students with knowledge and skills to apply multi-level models to their own research contexts.