Short Courses

July 1 1-4 PM CET

Alessandra Gaia: Indirect Questioning Techniques for Sensitive Topics

July 8 1-4 PM CET

Stefan Jünger: First Steps on Using R as a Geographic Information System

Katharina Meitinger, Cornelia Neuert, Dorothée Behr: Addressing comparability of measures before and after data collection

July 15 1-4 PM CET

Bugra Güngör: Designing an Online Survey

July 22 1-4 PM CET

Johannes Breuer, Stefan Jünger: Introduction to Survey Data Cleaning Using Tidyverse in R

Angelo Moretti: Introduction to Applied Small Area Estimation

Indirect Questioning Techniques for Sensitive Topics

July 1^st 1-4 pm CET

Instructor Bio:

Alessandra Gaia, PhD

Department of Sociology, University of Milan-Bicocca

Dr. Alessandra Gaia is a Research Fellow at the University of Milano-Bicocca. Previously she worked as a survey manager at the European Social Survey headquarters (City, University of London) and at the UCL Centre for Longitudinal Studies; also, she worked at the Institute for Social and Economic Research of the University of Essex. Her scientific papers are published, among others in the following journals: Methods, Data, Analysis; the Journal of Survey Statistics and Methodology; Social Science Computer Review, and other academic journals. Her research interests are: research methods for hard-to-reach populations, survey methodology, and methodology of social research.

Course description:

Researchers often need to collect data on topics which are sensitive for respondents, such as, for example, drug consumption, illegal behaviours, voting or income. In order to minimise social desirability bias and increase data quality in questions on sensitive topics, survey methodologists ideated a number of indirect questioning techniques. These are techniques that, rather than asking respondents’ directly to report the sensitive attribute, use randomisation processes or probabilistic theories to hide respondents’ answers and augment privacy. One of these techniques is the Item Count Technique (also known as unmatched count technique or list experiments).

This course provide a description of the Item Count Technique and its variations (i.e. the Item Count Technique, The Two Lists ICT, The Longitudinal Item Count Technique, The Item Sum Technique, The Person Count Technique), and offers examples of applications of the Item Count Technique in the literature. Students will also learn best practices to design and administer ICT questions in surveys. At the end of the course students will be able to: recognise the different techniques, understand which technique is more appropriate in each context, and design ICT questions. The course will be interactive in nature and students will be prompted to test different type of questions.

Using the smartphone application Wooclap students will be given the opportunity (if they wish) to answer a number of survey questions administered through the Item Count Technique and other indirect questioning techniques. This exercise will help students grasp the idea behind these methods and provide insights for discussion.

Requirements:

This is an introductory course, however basic knowledge in statistics and quantitative research methods would be beneficial.

First Steps on Using R as a Geographic Information System

July 8 1-4 PM CET

Instructor bio:

Stefan Jünger, PhD

GESIS – Leibniz Institute for the Social Sciences

https://stefanjuenger.github.io

Dr. Stefan Jünger studied sociology, philosophy, and cultural anthropology and pursued a Ph.D. from the University of Cologne in 2019. As a postdoctoral researcher and acting head of the GESIS Secure Data Center (SDC), he coordinates the emergence of Geographic Information Systems (GIS) in contemporary data management workflows and supervises the georeferencing of GESIS data. His research interests are social and ethnic environmental inequalities, attitudes towards minorities, data privacy, data documentation, and open science. He works squarely at the intersection of empirical social science research, GISciences, and research data management.

Course description:

When social scientists aim to use geospatial data, they must rely on specialized tools, called Geographic Information Systems (GIS). GIS are tools to visualize, process, and analyze geospatial data. However, the world of GIS is complicated since often only foreign software solutions provide a comprehensive collection of available geospatial techniques. Fortunately, nowadays, social scientists can also use the statistical software R as a proper GIS. In recent years, not only the range of supported geospatial file formats in R was expanded, it is nowadays way easier to wrangle such complicated data. Most importantly, the collection of applicable geospatial techniques heavily increased.

This course will teach the first steps of exploiting R and applying its geospatial techniques. Participants who are at least on an intermediate-level of expertise in R will learn about the most common data formats, how to get the data, and wrangle them for further analysis. Finally, many researchers’ interest is creating maps, which is also straightforward and fun to do in R. The course will therefore end with plotting maps to give participants a visual memento and stimulate future work with geospatial data in R.

Requirements:

Intermediate-level of expertise in R.

Participants need a working installation of R and RStudio.

Addressing comparability of measures before and after data collection

July 8 1-4 PM CET

Instructor bio’s:

Katharina Meitinger, PhD

Department of Methods & Statistics, Utrecht University

Cornelia Neuert, PhD

GESIS – Leibniz Institute for the Social Sciences

Dorothée Behr, PhD

GESIS – Leibniz Institute for the Social Sciences

Dr. Katharina Meitinger is an assistant professor in methods and statistics at Utrecht University. Before joining Utrecht University, she worked at GESIS Leibniz-Institute for the Social Sciences, since 2013 and at the University of Mannheim since 2016. In her research, she works on qualitative pretesting approaches, equivalence assessment of survey items, and mixed methods approaches. She received multiple international awards for her work on mixed methods approaches for the assessment of cross-national data (AAPOR/WAPOR Janet A. Harkness Award, NCHS Monroe Sirken Innovative Award for Young Scholars of Question Evaluation, ASI Young researcher award for excellent contributions to social science research).

Dr. Cornelia Neuert is a social scientist and head of the team Questionnaire Design & Evaluation at GESIS – Leibniz Institute for the Social Sciences. She has been working as research associate in the GESIS pretesting unit since April, 2012. Together with the staff of the pretesting unit, she has conducted numerous cognitive pretests for various research projects and survey programs. Her research focuses on methods for testing and evaluating survey questionnaires and questionnaire design.

Dr. Dorothée Behr is a cross-cultural survey methodologist and head of the team Cross-Cultural Survey Methods at GESIS – Leibniz Institute for the Social Sciences. She has a doctorate in applied translation studies (topic: questionnaire translation) from the University of Mainz. Her research and services (consultancy, training, and project involvement) focus on questionnaire translation as well as comparability of cross-national survey data, as assessed through web probing. Major large-scale projects include PIAAC Cycle 1 and 2, where she was responsible for the translation guidelines for the background questionnaire. She is also a member of the Translation Expert Panel of the ESS.

Course description:

Cross-national data production in social science research has increased substantially in recent decades. However, dealing with international instead of national data adds an “additional layer of complexity” to the development and analysis of such measures. Respondents come from different contexts, cultures, and speak different languages. Different sources of bias can threaten the comparability of data. A concept might not exist in one of the countries of the study (construct bias) or the translation of a key term might capture different nuances than the source question (item bias). Therefore, it is important to be aware of these sources of error before the actual data collection and take a systematic approach on how to keep them minimal. After data collection, it is crucial to systematically assess whether the measures are indeed comparable because measures might have still been biased.

This course will provide an overview of the process to develop and assess cross-national measures. After a short introduction of different sources of bias and the Total Survey Error in a cross-national context, this course will focus on three steps in the process:

1) How to achieve a comparable understanding of questions?

2) How to translate?

3) How to assess the comparability of measures after data collection?

The course will discuss crucial steps in the development of the questionnaires, such as input of cultural experts, cross-cultural cognitive interviewing, and translatability assessments. It will highlight important elements of the translation process, such as the TRAPD approach to translation. Finally, it will provide an overview of quantitative and qualitative approaches to evaluate the comparability of measures after data collection, such as measurement invariance tests and web probing.

The course will be a mixture of interactive lectures and short exercises.

Requirements:

None, beginner level.

Designing an Online Survey

July 15 1-4 PM CET

Instructor bio:

Bugra Güngör

Graduate Institute of International and Development Studies, Genève

Bugra Güngör is a PhD Candidate in International Relations and Political Science and Teaching Assistant in MINT programs at the Graduate Institute of International and Development Studies. Regarding survey methodology, he took a doctoral course titled “Survey Experiments”, he attended summer schools on “Web Survey Design” and “Designing, Implementing, and Analyzing Longitudinal Surveys” organized by the GESIS in 2018 and 2020. Lastly, as a part of his dissertation, he has conducted an online list experiment on the support for electoral violence in the most and least competitive districts of Istanbul prior to the June 2019 municipal elections.

Course description:

Online surveys have become an important part of understanding individuals’ attitudes and opinions on certain issues, thanks to its advantageous sides compared to face-to-face surveys like budgetary constraints, shorter completion period, elicitation of honest views, the inclusion of audio and visual elements and so on. Therefore, web surveys have been accepted as a significant tool through which the researchers explore public opinion from marketing studies to election polls. In particular, the instructor plans to run the following programme:

First hour:

Overview and focus of the short course
Pros and cons of online surveys
Design principles
Question designs

Second hour:

Use of interactive items and multimedia
Various ways of scaling
Scrolling and paging designs
Alignment and spacing

Third hour:

Different types of assistance and progress indicators
Handling dropouts and nonresponse
Wrap-up

The primary focus of the short course is to teach participants how to come up with the best possible survey instruments for web-based data collection. The course will cover the impacts of design on measurement error in online surveys; what is more, it will demonstrate practical examples in order to ensure maximum quality in data collection. Participants may bring their own projects prior to the course to receive feedback on their survey designs.

During the workshop, participants will be asked to identify various design problems in the light of examples the instructor shares with them. In addition, they will come up with some alternative ways to adjust the examples according to the basic design principles the instructor teaches.

Requirements:

None, beginner level.

Introduction to Survey Data Cleaning Using Tidyverse in R

July 22^nd 1-4 PM CET

Instructor bios:

Dr. Stefan Jünger

GESIS – Leibniz Institute for the Social Sciences

https://stefanjuenger.github.io

Johannes Breuer, PhD

GESIS – Leibniz Institute for the Social Sciences

https://www.johannesbreuer.com

Dr. Johannes Breuer works as a senior researcher at GESIS – Leibniz Institute for the Social Sciences where he is the acting head of the team Data Linking & Data Security. He received his Ph.D. in psychology from the University of Cologne in 2013. Before joining GESIS, he worked in several research projects investigating the use and effects of digital media at the universities of Cologne, Hohenheim, and Münster, and the Leibniz-Institute für Wissensmedien (Knowledge Media Research Center). His other research interests include computational methods, data management, and open science.

Course description:

Before researchers can start to analyze their data, they first have to wrangle (i.e., clean and transform) them. While it may not the most exciting part of data analysis, it can take up a substantial part of the researchers’ time. An often-used phrase applies the pareto principle to working with research data and states that 80% of the time is spent wrangling the data, and only 20% actually analyzing it. Most statistical software packages offer various options for data wrangling that differ in their accessibility and versatility. Among these options, the R programming language is a very powerful tool for data wrangling. While all data wrangling can be done with base R, the syntax for this is typically verbose and not intuitive and, hence, difficult to learn, remember, and read. The tidyverse, which is “an opinionated collection of R packages designed for data science” in which “all packages share an underlying design philosophy, grammar” see https://www.tidyverse.org/), addresses this problem by providing a consistent syntax that is also easy to read, learn, and remember. This workshop will introduce participants to the tidyverse and its packages as well as the concepts that it builds on, such as tidy data. In the workshop’s practical parts, we will work through examples of common data wrangling steps: importing, tidying, and transforming data.

Requirements:

Intermediate level & Advanced level.

The course is meant for people who already have some experience with R looking for an accessible, hands-on introduction to data cleaning with the tidyverse as well as more advanced R users who want to switch from base R to the tidyverse for their data cleaning tasks.

Participants will need a working installation of R and RStudio and should, ideally, also install the tidyverse packages before the course by running the command install.packages(“tidyverse”) in R/RStudio.

Introduction to Applied Small Area Estimation

July 22^nd 1-4 PM CET

Instructor bio:

Angelo Moretti, PhD

Department of Computing and Mathematics, Manchester Metropolitan University Dr. Angelo Moretti is a lecturer at Manchester Metropolitan University, holding a PhD in Social Statistics from the University of Manchester. His research interests lie in survey statistics, in particular small area estimation of social indicators, data integration and survey calibration. He is also working on composite social indicators and data dimensionality reduction problems in small area estimation. In addition, his research focuses on applications in different domains such as, wellbeing, poverty and crime.

Course description:

Large-scale sample surveys are not designed to produce reliable estimates for small population

domains, e.g., geographical areas or population groups. Therefore, small area estimation methods, that borrow strength information from auxiliary data e.g., the Census or administrative data, can be used to produce reliable estimates. This course covers basic small area estimation methods based on the direct and model-based estimation approach and it is structured in three parts. The first part is about the introduction to the small area estimation problem and the use of direct estimators to produce small area estimates. In the second part, we introduce the unit-level approach based on the Battese, Harter and Fuller model, assuming that auxiliary information is available at unit-level. The third part is on the area-level approach, based on the Fay-Herriot model. This approach is useful when the auxiliary information is available are area-level only. Applications and examples in R are presented in each part.

By the end of the course participants will be able to:

Understand the small area estimation problem
Understand which techniques are most commonly used (and why)
Be able to apply and validate two of the most diffused small area estimation methods based on the area-level and unit-level approach
Implement the methods in R software.

During the course participants will be provided with the R programs and datasets needed to produce the analysis presented in the course. The intention of this course is providing useful applicable tools for researchers and practitioners. It will be a mixture of different activities, i.e., methods, practical examples and software applications.

Requirements:

None, beginner level.

Ideally, participants have a working installation of R and RStudio.