Rami Tabri, University of Sydney - Robust and Credible Comparisons of Distributions Under Missing Data

Title: Robust and Credible Comparisons of Distributions Under Missing Data

Abstract:  Nonresponse is a universal problem in all voluntary surveys, and credibly accounting for it is critical for sound empirical practice. The presentation's focus is on the comparison of distributions and the presentation of new tools that can aid/empower empirical researchers in addressing the missing data problem in this context. The general approach is to calculate distributional bounds that are compatible with the observed data and the maintained assumption on nonresponse. While such bounds may be wide in practice, we show that they can be informative for empirical analysis and may yield conclusions that are robust to the nature of nonresponse. The presentation is nontechnical and illustrates the tools with two empirical applications: (i) comparison of trust levels in Lebanese public institutions, and (ii) comparisons of income distributions using the HILDA Survey.

Presenter: Rami Tabri, University of Sydney

If you would like to subscribe to the Melbourne Institute Seminar Series email list, please contact us.