Statistical Tools for Analysis of Nonprobability Samples

Music City Center 

Room: CC-107B 

The use of non-probability samples for descriptive inference is becoming increasingly popular, though statistical tools for this purpose are somewhat limited. Developing statistical tools for analyzing data from non-probability samples is an important practical problem in survey sampling. The topic is also quite related to handling missing data.
In this short course, we will cover state-of-the-art tools for analyzing non-probability samples. The primary statistical tool we will focus on is propensity score weighting, which addresses selection bias in the sample. We will rigorously cover statistical theory and methods for propensity score weights. We also cover how to perform statistical inference using these weights. Additionally, we will introduce relevant R packages and give a demonstration using real-life data examples.

Prerequisite: Basic knowledge (undergraduate level) in mathematical statistics and survey sampling. Experience in using R software.

Survey Research Methods Section