Course 9: Sampling, Weighting, and Estimation

Lecturer: Stephanie Eckman

Date: 20-24 August 2018
Time: 09:00-13:00, 14:00-16:00

Short Course Description:

This course will cover: methods of sample selection; calculation of weights that adjust for nonresponse and undercoverage; and analysis of complex survey data. We will also discuss analysis of nonprobability surveys. The emphasis of the course is more applied than theoretical, but students are expected to be comfortable with statistics and to have some experience with data analysis. For each topic, students will do exercises in Stata that apply the techniques learned in the lectures. Students will get the most out of the class if they have prior experience with Stata.

Course Prerequisites:

  • Introductory course in statistics. No prior knowledge of sampling theory is assumed, but students should be comfortable with statistical concepts such as hypothesis testing, variance, standard errors, confidence intervals, etc.;
  • Prior knowledge of Stata is required for this course;
  • Basic understanding of survey methodology (this could be gained in the course “Introduction to Survey Design” in the first week);
  • Experience in handling survey data is helpful but not necessary.

Target Group:

Participants will find the course useful if:

  • they have some experience conducting surveys and/or analyzing social science data, but have not yet studied sampling;
  • they are conducting their own survey or are analyzing survey data.

Course and Learning Objectives:

By the end of the course participants will

  • have a sound understanding of the most frequently used sample designs (one- and two-stage sampling, clustered sampling; stratified sampling, and related designs);
  • know how to create probability weights, and have experience with several methods for adjusting such weights for nonresponse and undercoverage;
  • understand how and why the design of a sample survey affects the analysis of the data;
  • know the appropriate methods to use in Stata to analyze complex survey data, and the pros and cons of each method.

Organizational Structure of the Course:

The course will consist of 4 hours of lecture in the morning. Students will spend the afternoons working on exercises (most of them in Stata) designed to deepen their understanding of the material. The instructor will be available in the afternoons to answer questions.

Software and Hardware Requirements:

None. This course takes place in a computer lab. GESIS will provide participants with access to the statistical software package used in the course (i.e., Stata) as well as other packages (e.g., SPSS, Mplus, R). Access to an own laptop computer might turn out to be useful.