Design and implementation of a high-quality probability sample: Lessons learnt

Migration research is one of the most rapidly developing fields of demography in Europe. To develop an empirically-based theoretical understanding of immigration we need high-quality representative data. However, surveys of immigrants often are confronted with challenges of coverage, representativeness and response rates and cannot face the high costs needed to overcome them. It has been argued that, when registered data is not available, the highest quality approach is probability sampling with household screening.

surveyThe aim of this paper is to report on a high-quality survey of this kind: The UK Household Longitudinal Study (UKHLS). The survey started in 2009 with an overrepresentation of ethnic minorities. Because of the shrinking of some groups due to sample attrition and in order to represent the households that entered after 2009, it was decided to introduce a new sample that would cover ethnic minority groups and recent immigrants- the Immigrant and Ethnic Minority Boost (IEMB). The challenges related to designing a representative sample concerned the trade-off between cost and coverage, and the declining response rates, particularly across minorities and immigrants. To overcome these challenges, a small-area census data was used to identify areas that covered the majority of immigrants and target ethnic minority populations and  addresses from those areas were oversampled using varying sampling fractions. Households were screened for eligibility based on whether they included a target immigrant/ethnic minority member. If so, all adult members were interviewed.

Overall, although the IEMB boost has not resulted in a fully representative sample, the UKHLS sample provides large numbers for analysis on several dimensions and is broadly reflective of the 2011 Census. The IEMB provides substantial samples for analysis of variation across national, ethnic, and regional origins and of generational change. The paper therefore suggest that a high-quality approach based on probability sampling and population screening offer a resource to the rapidly evolving field of migrant and ethnic minority studies.

The full article can be accessed on Demographic Research

Photo credit to Ikhlasul Amal