S030: Intermediate Statistics for Educational Research: Applied Linear Regression

The HOW: The S-030 experience engages us in building a supportive course community where we all learn from one another. In place of more traditional pre-work such as reading chapters or articles, we will instead use videos (filmed by the instructor specifically for this course) and interactive opportunities for practice as preparation for each class meeting. This will allow us to use our time together to explore statistical topics in a deeply engaged and participatory format. Throughout the course, we will focus not only on the methods themselves but on why and how their use can represent a powerful tool for enacting change. We will emphasize the summarizing and presentation of statistical results for multiple audiences. Our focus will be on understanding how to use and interpret linear regression modeling to become both critical consumers and ethical producers of quantitative work. 

The WHAT: This course goes in-depth through a set of quantitative methods referred to as the general linear model -- regression, correlation, analysis of variance -- to address various questions in education, psychology, and social science research. Examples of research questions we will be addressing are: How do children's demonstrations of separation distress relate to their age? Do mandatory seat belt laws save lives? Which animal has the most massive brain, relative to their body mass? and many more. We will learn statistical analysis by doing statistical analysis. Building on the basic principles of elementary statistics (the content of S-012 or equivalent), we will begin with the simple linear regression model. We'll then utilize its mechanisms to fit models that address increasingly sophisticated research questions. As we progress, we'll learn to fit models that can accommodate: multiple predictors; transformations to address non-linearity; dichotomous and polychotomous predictors (both ordinal and nominal); statistical interactions; and quadratic relationships. We'll include a heavy emphasis on the substantive interpretation and communication of results. We will use Stata as our statistical software in this course. There will be multiple assignments -- some to be completed individually, some with a partner -- where students will address a variety of research questions through scaffolded analysis of real datasets, as well as a final project presentation.

The MONDAY SECTION: While the whole class meets together on Tuesday mornings, students are also expected to attend one 75-minute smaller-group section led by a Teaching Fellow on Mondays. Final assignment of section times will not be done on my.harvard during enrollment, but rather will be coordinated by the teaching team after enrollment has occurred.