Alessandra Gaia (Ph.D.) is a researcher at the University of Milan-Bicocca. Previously she worked as a survey manager at the European Social Survey Headquarters at City, University of London; she worked as survey researcher at the Institute for Social and Economic Research of the University of Essex, where she was part of the team that administers Understanding Society: the UK Household Longitudinal Study. She has also served the UCL Centre for Longitudinal Studies as survey manager of Next Steps: the Longitudinal Study of Young People in England. She briefly collaborated with the United Nation Department of Economic and Social Affairs and with the World Food Programme of the United Nations. Her scientific papers are published in the following journals: the Journal of Survey Statistics and Methodology; Methods, Data, Analysis; Survey Methods: Insights from the Field, and other academic journals. She is currently contributing to a large encyclopaedia of social sciences, the SAGE Research Methods Foundations: An Encyclopaedia. Her research interests are: research methods for hard-to-reach populations, data quality when asking sensitive questions in surveys, social desirability bias, and Methodology of Social Research.

Course Content

The course presents quantitative research methods to conduct research on rare, marginal, hidden, and elusive populations (also called hard-to-reach populations), such as, for example, sex workers, illegal immigrants, victims of trafficking, and drug users. After introducing and defining hard-to-reach populations, the course describes a wide range of methods: techniques to estimate the size of hard-to-reach populations (capture-recapture), sampling strategies (Respondent Driven Sampling, RDS), and data collection methods to ask questions about sensitive topics, including indirect questioning techniques (e.g. the Item Count Technique, the Randomised Response Technique, etc). Ethical issues arising when investigating hard-to-reach populations will be discussed. The course is applied in nature and includes examples from empirical research.


Course Objectives

The course is addressed to Ph.D. and Master students, academic researchers, and practitioners (from NGOs, UN, International Organizations, charities, and survey agencies), that wish to conduct research on hard-to-reach populations. Participants will learn how to sample and collect data on rare, marginal, hidden, elusive and excluded populations, in developing countries as well as in the western world. Also, students will become familiar with the ethical issues associated with conducting research in these contexts. Participants will be encouraged to present their work and the instructor will discuss case studies taking into consideration students’ areas of research interest.


Course Prerequisites

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


Representative Background Reading

Kish, L. (1991). A Taxonomy of elusive populations. Journal of Official Statistics, 7(3): 339-347.

Tourangeau, R. (2014). “Defining hard-to-survey populations”. Hard-to-Survey Populations. Cambridge: Cambridge University Press. doi:10.1017/CBO9781139381635, p: 3-20.


Required texts

Tourangeau, R., Edwards, B., Johnson, T., Wolter, K., & Bates, N. (Eds.). (2014). Hard-to-Survey Populations. Cambridge: Cambridge University Press. doi:10.1017/CBO9781139381635



Day 1:

Session 1: Introduction and key concepts

Introduction to the course. What are hard-to-survey populations? What is the difference between rare, marginal, excluded, elusive, and hidden populations? What are the challenges experienced by social researchers in studying hard-to-survey populations? What is a “sensitive question”? What is social desirability bias and how does it affect data quality?

Session 2: Measuring undercounts and estimating the size of hard-to-survey populations

Administrative record match; demographic analysis; reverse record check; post-enumeration survey. Estimating the size of hard-to-survey populations: the capture-recapture method. Case study: the prevalence of crack cocaine users in London.

Day 2:

Session 3: Sampling hard-to-survey populations (part 1)

Introduction to sampling methods for hard to survey populations: screening, venue-based and time-location sampling, random walks, indirect sampling.

Session 4: Sampling hard-to-survey populations (part 2)

Introduction to network-based sampling methods: snowball sampling and Respondent Driven Sampling (RDS). Case study: sampling men who have sex with men.

Day 3:

Session 5: Data collection on hard-to-survey populations

Asking sensitive questions in surveys; mode effects; interviewer effects; privacy of the interview setting; forgiving introduction; direct versus indirect questioning techniques. Case study: mode effects and the sexual double standard.

Session 6: Indirect questioning techniques (part 1)

The Randomised Response Technique; the nonrandomized response technique (NRRT); network scale-up technique (NST); the three card method.

Day 4:

Session 7: Indirect questioning techniques (part 2)

The Item Count Technique(list experiments); the “two lists” version of the Item Count Technique; the Longitudinal Item Count Technique; the Person Count Technique; the Item Sum Technique.

Session 8: Indirect questioning techniques (part 3)

Designing the Item Count Technique; analysing data from the Item Count Technique. Case study: an application of the Item Count Technique to study vote buying in Nicaragua.

Day 5:

Session 9: Ethics

The ethics of surveying hard-to-reach populations; data privacy and anonymisation; panel conditioning and the question behaviour effect.

Session 10: Conclusion

Discussion of students’ assignments; Summary; Q&A.