S-052 Applied Data Analysis
John B. Willett
S-052 is designed for those who want to extend their data-analytic skills beyond a basic knowledge of multiple regression analysis, and who want to communicate their findings clearly to audiences of researchers, scholars, and policymakers. The course contributes directly to the diverse data-analytic toolkit that the well-equipped empirical researcher must possess in order to perform sensible analyses of complex educational, psychological, and social data. Topics in the course include more extensive use of transformations in regression analysis, influence statistics, building and comparing taxonomies of regression models, general linear hypothesis testing, an introduction to multilevel modeling, nonlinear regression analysis, binomial and multinomial logistic regression analysis, principal components analysis, cluster analysis, introduction to discrete-time survival analysis, dealing with missing data, and others. S-052 is an applied course that offers conceptual explanations of statistical techniques, along with opportunities to examine, implement, and practice them in real data. Learning the computer skills necessary to conduct these kinds of analyses, and the communication skills to discuss them, is an integral part of the course.
Visit the course Web site
(Some resources on the course Web site may require a Harvard PIN number)
Prerequisite: Successful completion of S-030 or permission of the instructor.