Applied Survey Data Analysis

Date: 
Saturday, March 30, 2013 - 09:15
Location: 
Business and Economics Bldg, Rm. 313-BE

Organizers

  • Applied Statistics Lab (ASL)
  • Quantitative Initiative for Policy and Social Research (QIPSR)
  • Department of Political Science

Overview

Why do we need survey data analysis software? Most regular statistical software analyzes survey data as if the data were collected using simple random sampling, which is rarely the case. When surveys are stratified, clustered or weighted, we need special statistical software to estimate standard errors accurately and avoid a host of estimation errors.

This workshop provided participants with an introductory, hands-on overview of issues frequently encountered when analyzing survey data. The first part of the workshop reviews different sampling techniques, core statistical concepts, and methods of variance estimation. In the second part, participants receive hands-on experience in the analysis of survey data using the “svy” commands in Stata. Examples are drawn from the analysis of social science surveys such as the American National Elections Studies (ANES), General Social Survey (GSS), and AmericasBarometer surveys.

 

Program

Introduction

Mark Peffley, Professor, Department of Political Science and Director and QIPSR

Arnold Stromberg, Chair, Department of Statistics and Co-Director, Applied  Statistics Lab (ASL)

Introduction to Sampling Techniques: Simple Random Samples versus Complex Samples

Abby Córdova, Assistant Professor, Department of Political Science  

Dominique Zéphyr, Data Analyst, Applied Statistics Lab (ASL)

Variance Estimation Techniques

Arnold Stromberg, Chair, Department of Statistics and Co-Director, Applied  Statistics Lab (ASL)

Analysis of Complex Survey Data using StataDo File, and Link to Data

 Dominique Zéphyr, Data Analyst, Applied Statistics Lab (ASL)

 

Materials: 
Event type: 
Statistical Workshop
X
Enter your Quantitative Initiative for Policy and Social Research username.
Enter the password that accompanies your username.
Loading