Using an Onomastic Approach to Gain Insights from Migrant Groups? Lessons from the Social Survey Austria 2016

The current political climate makes research on immigrants, their position in society, their values and attitudes essential. This includes second and third generation immigrants, as their input helps to understand their perceived position and experiences in host societies.

But they are only marginal parts of populations and representative population surveys only provide limited information on them, that is often unsuitable for statistical analysis.

Additionally, specialized research is difficult, as representative samples on such populations are hard to get a hold of. In most cases there are no adequate sampling frames, language barriers exist and interviewers have a hard time to gain access to the field.

Facing those challenges, the researchers of Social Survey Austria 2016 (CAPI; n=2000), decided to conduct an additional CATI, exclusively focusing on two sizeable migrant populations in Austria – Turks or Serbo-Croatians.

For the CATI the interview partners were recruited via an onomastic approach, searching for last names and that may indicate Turkish or Serbo-Croatian origins. Once identified, the interviews were offered in three languages and conducted by native speakers.

The resulting two n=300 samples provided more diverse and detailed data on the migrant populations, than the 82 Serbo-Croatians and 47 Turks in the main survey, especially regarding value orientation and societal standing. This will be illustrated by highlighting key dimensions in a comparison between the migrants in the main survey and the migrant samples.

Yet several other issues could not be addressed: Only a hand full of third generation migrants could be recruited, a large percentage of neutral dropouts was reported and bias towards larger cities could be observed.

Based on this, we illustrate that the onomastic approach provided a suitable approach to recruit people for these hard to reach parts of the population, but may only be used in conjunction with a general population survey.