2.7 Sample Validation

We calculated the participation rates in the different locations of interviews. To validate the DiPAS sample for representativeness, we compared it to several existing statistics on the population of refugees/asylum seekers in Austria and Vienna, in terms of numbers, age, and citizenship.

Participation rates in the different housing locations were calculated for the locations where the numbers of adult residents were provided. In the largest emergency quarters participation rates amount to 28% and 44% respectively, and in two of the smaller quarters they were as high as 64% and 77%. Unfortunately, participation rates cannot be calculated for all seven locations due to the incompleteness of provided lists or data confidentiality issues in some locations.

The DiPAS sample is first compared with the number of asylum seekers in Austria. The size of the DiPAS sample corresponds to 1.2% of persons seeking asylum in Austria in 2015. Since all of the respondents arrived before December 2015, the size of the sample increases to 1.4% of those arriving in the first eleven months. Available aggregated data allow a differentiation by citizenship and broad age groups. A comparison of all asylum seekers in Austria in 2015 (excluding unaccompanied minors) with the 972 asylum-seeking individuals living in Austria captured in the sample reveals an almost identical distribution in terms of age.

Further specification by citizenship shows that among Afghans, the proportion of children included in the DiPAS sample is higher than in the general refugee population (41% versus 35%), which might be explained by the fact that the comparably small group of Afghan respondents includes several persons with particularly large families (4 to 6 children). Among Syrian asylum seekers captured in the DiPAS sample, the share of children below the age of 18 is smaller (25% versus 34%), whereas more Syrians were aged 18 to 45 years (65% versus 59%) or 46 to 60 years (9% versus 6%). Deviations in the distribution by large age groups among Iraqis are very small and for those with other citizenships the distribution in the DiPAS sample is almost identical to the one for the whole of Austria. To summarise, the comparison with the total of roughly 80,000 asylum seekers (not including unaccompanied minors) by age group and citizenship indicated no evidence of a substantial bias in the DiPAS sample.

Next, our sample was compared with available official data on individuals residing in emergency quarters and receiving basic assistance in Vienna. The sample (including respondents, their spouse, and children if living in Austria) represents 15% of individuals residing in emergency quarters in Vienna at the time of the fieldwork. The proportion is high for Syrian and Iraqis (19% and 28% respectively), meaning that the current survey captures a substantial share of Iraqi and Syrian asylum seekers living in Vienna. However, the proportions of Afghans and other citizenships are low. Given this low share on the one hand, and the overall low number of Afghans in the DiPAS sample on the other hand, the results for Afghan asylum seekers should be considered with caution. Nevertheless, they provide valuable insights into this population.