Survey Research: Statistical Analysis and Estimation

About

The course is based on a total survey error perspective and discusses the major sources of survey error. Participants will be presented with tools for detection and adjustment of such errors. Analysis methods are introduced using both SPSS and R. Topics include complex sampling, nonresponse adjustment, measurement error, analysis of incomplete data and advanced use of administrative data. Special attention will be given to the analysis of complex surveys that include weighting, stratification and design effects.

Changes in technology and society strongly influence modern survey research. This course covers the essentials of modern survey analysis and estimation and is organised by the Department of Methodology and Statistics (UU) in collaboration with Statistics Netherlands (CBS). Lectures, practical applications, and computer classes are alternated. The course is intended for advanced students and professionals in such fields as social and behavioral research, marketing, business, health sciences, and official statistics. The course aims at researchers who intend to design and analyse their own survey, but also at researchers who analyse secondary data sets, such as the European Social Survey (ESS) or the International Social Survey Program (ISSP).

Central to the course is survey quality and the reduction of Total Survey Error (coverage, sampling, nonresponse, adjustment, measurement error, and processing error). Participants will be presented with tools for detection and adjustment of such errors. Analysis methods are introduced using both SPSS and R. Lectures and computer classes cover basic ideas from the TSE-perspective, sampling and non-sampling error: an introduction in R, survey estimation and inference, complex sampling, nonresponse adjustment, and analysis of incomplete data. Special attention will be given to the analysis of complex surveys - including weighting, stratification, and design effects - and to administrative data.

This course assumes knowledge of survey methodology and statistics. Participants should be acquainted with Analysis of Variance, Multiple Regression Analysis, standard errors, and have some hands-on experience with SPSS. No prior knowledge of R is assumed.

A good preparation is our Summer School course ‘Survey Research: Design, Implementation and Data Processing’ (S15, 16 -20 August).

The course is taught from ‘De Leeuw, Hox & Dillman (2008). International Handbook of Survey Methodology. New York: Taylor & Francis’. The book is not included in the course materials. This book has to be purchased in advance. (e.g. Bol.com, Amazon.de).

Please note that there is always the possibility that we have to change the course pending COVID19-related developments. The exact details, including a day-to-day program, will be communicated 6 weeks prior to the start of the course.

Course leader

Dr. Bella Struminskaya & Prof. dr. Barry Schouten

Target group

The course is intended for advanced students and professionals in such fields as social and behavioral research, marketing, business, health sciences, and official statistics. The course aims at researchers who intend to design and analyze their own survey, but also at researchers who analyze secondary data such as the European Social Survey (ESS) or the International Social Survey Program (ISSP). This course assumes general knowledge of survey methodology and statistics. Participants should be acquainted with the basics of Analysis of Variance, Multiple Regression Analysis, standard errors, and have some hands-on experience with a statistical package (e.g., SPSS, Stata, SAS). No prior knowledge of R is assumed. A maximum of 40 participants will be allowed in this course. Please note that the selection for this course will be done on a first-come-first-served basis.

Course aim

This course aims to provide participants with state of the art knowledge and application oriented skills for survey analysis and estimation. After the course, participants are ready to apply the learned towards their own data or archived data sets, and are able to take advanced training in complex survey statistics and adjustment.

Fee info