Using STATA for Survey Data Analysis

June 12, 9:00 am – 12:00 pm, 1:30 – 4:00 pm
Stanislav Kolenikov, University of Missouri
 

This workshop is intended for survey statisticians who want to familiarize themselves with survey capabilities of Stata software. The workshop will match the methodological concepts in survey sampling and estimation with appropriate Stata commands, proceeding from descriptive means, totals and proportions to regression and GLM estimation commands. The main variance estimation methods (Taylor series linearization, jackknife, BRR and the bootstrap) will be introduced and exemplified. The bootstrap methods will be discussed in particular detail.

The attendees must be familiar with the basic concepts of sampling (stratification, clustering, varying probabilities of selection, use of sampling weights) and design-based estimation (weighted estimation, Horvitz-Thompson estimator, Taylor series linearization for variance estimation). Familiarity with resampling variance estimators (BRR, jackknife) is an advantage. A basic knowledge of Stata is presumed.

To maximise their learning experience, participants are encouraged to bring their own computer with Stata installed on it. Note that bringing a computer will however not be essential to follow the workshop.