Welcome to the CoursePlus Web site for METHODS IN ANALYSIS OF LARGE POPULATION SURVEYS (380.712.01), a course offered by the Department of Population, Family and Reproductive Health at the Johns Hopkins Bloomberg School of Public Health.
Introduces the practical aspects of design and analysis of large sample surveys. Covers statistical issues of complex surveys involving stratification and clustering, methods of handling missing data, weighting, sample size estimation and allocation, design-based analysis of frequency tables and multivariate methods for complex surveys. Emphasizes applied, rather than theoretical derivation.
Upon successfully completing this course, students will be able to:
- Take survey design into consideration during analysis and perform design-based analyses of data from complex surveys, such as multistage national surveys
- Estimate variances with Taylor linearizations, jacknife, and bootstrapping methods in univariate and multivariate statistics
- Explain advantages and disadvantages and apply design weights
- Examine data missingness patterns and use appropriate imputation methods for missing data
- Compare and contrast design-based analyses to multilevel and marginal models for addressing intraclass correlation and design-effects
Mon 3:30 PM to 5:20 PM
140.640 or consent of instructor
Grading Restrictions: Letter grade