Relevance & Research Question:

The sampling of migrant populations is a highly complicated issue and one of the most pressing methodological challenges in migration research. Especially in cross-national studies this problem is intensified by the need of comparable sampling frames in different countries. Name-based telephone (i.e., onomastic) sampling is a well-established procedure in such settings. However, due to the declining use of landline connections and missing entries in telephone books the need to amend this technique by other approaches becomes increasingly visible. We therefore investigate whether and how a social networking site can be used to sample a migrant population in a cross-national setting. More specifically we ask whether the advertisement functionality of Facebook can be used for this purpose.

Methods & Data:

We build on results of an online survey amongst Polish migrants in four different countries (Austria, Ireland, Switzerland, and the United Kingdom) conducted in late 2015. Respondents for this study were sampled through Facebook advertisements. Since we want to assess the effectiveness of our approach, our data consists of the statistical information regarding our advertisement campaign provided by Facebook and of the data collected through our survey.

Results:

Using a total advertisement budget of €500 we succeeded in recruiting more than 1,000 respondents who completed our survey within a four weeks field period. These results are also remarkable as we used a relatively comprehensive questionnaire (median completion time 11.2 minutes). Our analysis shows that the approach is not only efficient but also accurate as nearly all of the respondents (96%) belong to the target population.

Added Value:

The study shows how Facebook advertisements can be used to sample migrants and hence counteract some of the most pressing problems of more established methods. Hence, we judge it to be a very valuable addition, yet in most scenarios not a replacement, of the latter.