Universitat Pompeu Fabra (Campus de la Ciutadella)

Barcelona

7-11 July 2014

The language of instruction for the courses will be English.

Deadline for registration is June 20, 2014.

Brief information on the courses is outlined below; for complete information, including the courses' programs, costs, and registration procedures, please download the following PDF program.

OBJECTIVES
Provide the participants the knowledge about the main issues related to the implementation of survey studies and the best way to deal with them.
Understand and apply sophisticated statistical techniques to survey data.

TARGET GROUP:
Business professionals who develop their work in survey methodology and/or statistical methods, teachers, researchers and students.

PROGRAM

Survey nonresponse
Instructor: Ineke Stoop, European Social Survey, The Netherlands Institute for Social Research/SCP

Nonresponse is a major concern of survey sponsors, survey agencies, and data users. Because of the decreasing response rates in many European countries survey costs increase and fieldwork periods lengthen, survey agencies have to enhance efforts to reach target persons and obtain their participation, and data users worry about the representativeness of the outcomes of their analyses and the accuracy of their estimates.

Surveys researchers worldwide work on improving survey designs, experiment with modes and incentives, and investigate how balanced response rates can be achieved, i.e. equal response rates from men and women, the rich and the poor, and people who like or dislike politics. They also try to collect auxiliary variables that both correlate with key outcomes of the survey and response propensities, to assess the presence of nonresponse bias and to adjust for this, if necessary and possible.

The course will present the nonresponse problem from a general survey quality perspective. It will delve into causes and correlates of nonresponse, describe measures to enhance response rates, such as interviewer training, incentives and advance letters, show how nonresponse bias can be assessed, and present some ways to adjust for nonresponse. Special attention will be paid to survey design aspects that exclude specific groups (e.g., the illiterate), the use of mixed mode designs, and nonresponse in a comparative perspective. The course will focus on nonresponse on surveys among individuals and households, rather than businesses. The emphasis is on unit nonresponse rather than item nonresponse.

The course will be useful for those who conduct their own surveys, who wish to evaluate the quality of data collected by a survey, and who wish to assess the possible effects of nonresponse on their analyses. A hands-on approach will be used, which means that input from the participants is planned in every phase of the course.

Some general knowledge on survey methodology is required.

Designing and Conducting Business Surveys.
Instructors: Jacqui Jones, Deputy Director of the Business Indicators and Balance of Payments Division in the UK Office of National Statistics (ONS) and Diane Willimack,  Chief of the Response Improvement Research Staff at the Economic Programs Directorate of the U.S. Census Bureau

Business surveys differ in important ways from social surveys, for example:

·Business entities are dynamic.

·Target populations are often skewed requiring certainty selection of large businesses in survey samples.

·Businesses are likely to be in more than one survey and in recurring surveys to support official statistics.

·The response process is complex and often involves more than one person.

·Business records may provide a source for requested survey data.

·Surveys rely heavily on self-completion data collection modes.

·Businesses can be classified into industrial classifications based on economic rules, which may seem arbitrary from the business perspective.

·Businesses may be re-contacted post-collection, e.g. during editing, to clarify reported data.

Because of these and other differences, practical issues emerge that have implications for survey design decisions at all stages in the survey process.

Using a process-quality perspective, derived from the Generic Statistical Business Process Model (GSBPM) and grounded in the Total Survey Error framework (Groves et al., 2009), this course provides an overview of methodological issues associated with the use of surveys to collect data from businesses.  We will:

Identify key differences between household surveys and business surveys, emphasizing organizational behaviors and attributes that affect survey response.

Demonstrate an approach to survey planning and design that utilizes understanding and consideration of this business context when developing, adapting, and implementing data collection instruments and procedures. 

Look at process quality measures throughout the business survey process to help effectively monitor and manage surveys.

This course will also include topics related to survey communication and response improvement strategies, managing and monitoring data collection processes, along with post-collection procedures such as editing, analysis and dissemination.

This integrated approach to surveys of businesses is the subject of the 2013 book published in the Wiley Series in Survey Methodology, entitled Designing and Conducting Business Surveys, written by Ger Snijkers, Gustav Haraldsen, Jacqui Jones, and Diane K. Willimack.

Spatial data analysis.
Instructors: Albert Esteve, Research Scientist and Deputy Director, Centre d’Estudis Demogràfics (CED)and Antonio López Gay, Research Scientist (CED)

This course provides an introduction to spatial data analysis for social scientists. The increasing availability of spatial data (both at the aggregated and at the individual level) has expanded the range of methodological tools to explore the spatial dimension of social phenomena and explain variation among areas. In this course, the student will learn how to manipulate and analyze spatial data. We will present basic techniques of Exploratory Spatial Data Analysis (ESDA), the concepts of local and global Spatial Autocorrelation, and introduce spatial (Lag and Error) regression models. Theoretical explanations will be accompanied by three lab exercises using GEODA (freeware). GEODA is a user-friendly software developed by Luc Anselin, the leading scholar in spatial econometrics, which enables beginners to quickly immerse in the world of spatial analysis.

Multilevel Modelling
Instructor: Leonardo Grilli, Associate Professor in Statistics , University of Florence

The course introduces the concepts of multilevel analysis, whose main aim is to model the relationships between and within groups. Typical situations include individuals clustered into families, schools, firms, geographical areas. The course focuses on the two-level linear model as a template to illustrate issues of specification, estimation and inference. The main ideas are illustrated by means of a couple of case studies. The second part of the course is devoted to special topics, such as the design effect and the required sample size, and to extensions, such as the logistic multilevel model for binary responses.

FEE FOR EACH COURSE: 
400€ professionals

300€ teachers and researchers

150€ students

A certificate of attendance will be provided.

INFORMATION AND REGISTRATION:

Research and Expertise Centre for Survey Methodology - RECSM
Universitat Pompeu Fabra, Barcelona

www.upf.edu/survey
recsm@upf.edu