8. Non-response among immigrants in Denmark

The politics of diversity: Social and political integration of immigrants in Iceland

“…The total response rate was 39%, but the response rates of the target groups with a foreign background were lower: a 14% response rate for individuals with a Lithuanian background, 11% for the Polish sample, and 14% for the Filipino sample. These low response rates are in line with trends visible in other countries, such as Denmark (Deding, Fridberg & Jakobsen 2008) the Netherlands (Feskens et al 2007;Feskens et al 2006;Dale & Haraldsen 2000), and Finland (Larja & Kotilainen 2014), where low response rates among immigrant groups have been identified as problematic, and researchers have argued that surveys must be tailored to fit immigrant groups. Although we acknowledge the problem of the low response rate among our participants with a foreign background, which limits the generalization capacity of the research, we nevertheless believe that because the topic is under-researched, it is important to proceed, to get the findings in the open, and to increase discussion on the position of people with a non-Icelandic background in Icelandic society.…”

Using population registers for migration and integration research: examples from Denmark and Sweden

“…First, as there are few incentives to report out-migration, some addresses randomly selected might not correspond to the real location of individuals. It is likely that this invalid address problem is more pronounced for cetain groups, thereby biasing the sample (Deding et al, 2008 ). Second, immigrants may become suspicious when information which they provide to state authorities (ie.…”

“…Meta-information from existing immigrant surveys conducted in the two countries (Groenewold & Lessard-Phillips, 2012 ; UIM, 2013 , 2014 , 2016 , 2017 ), as well as analyses of non-responses (Deding, Fridberg, & Jakobsen, 2008 ), illustrate that, even if immigrant surveys start with samples extracted from population registers, the final samples still suffer from pronounced non-reponse rates, which are much higher among immigrants than natives. Moreover, the various immigrant groups have different response rates: for example, the UIM surveys consistently report that immigrants over 30 years old have higher response rates compared to immigrants between aged between 18 and 29 years old, but both are significantly lower than the response rates of natives (between 40 and 50%, compared to ca.…”

“…Deding et al ( 2008 ) found that language problems do not drive the high non-response rate among immigrants, but that immigrant groups behave differently in survey situations: Pakistanis have the highest non-contact rate compared to Turks and Iranians (due to their having moved and their failure to meet the interviewers), while Turks have the highest refusals rate (driven by refusals by the spouse-husband). Albeit not at high rates, refusals due to suspicion occur more among immigrants than among natives (Deding et al, 2008 ). The TIES study reports similar contact and refusal problems for second-generation Turks in Sweden, which prompted the research team to draw a second sample from the register (Groenewold & Lessard-Phillips, 2012 ).…”

Settlement experiences of Iranian migrants on humanitarian and non-humanitarian visas in Australia

“…Scholars have used various methods to draw data from migrant populations. Surveys have been used extensively in various contexts in immigrant and refugee studies including Australia (Tilbury, 2006), the UK (Bloch, 1999;Erens, 2013), the USA (Smith, 2013), Spain (Duque, Baliano, & Perez, 2013;Mendez, Ferreras, & Cuest, 2013;Rinken, 2013), the Netherlands (Heelsum, 2013), Sweden (Myrberg, 2013), Denmark (Deding, Fridberg, & Jakobsen, 2013), and Switzerland (Lipps, Lagana, Pollien, & Gianettoni, 2013). Longitudinal surveys have also been used extensively in Australia (LSIA1, LSIA2, and LSIA3) (DIAC, 2012) to draw data from migrants as "surveys tend to be a versatile means of data collection in terms of both the arrays of topics that can be covered and adaptability of questionnaire and sample designs to address a particular research question in a particular setting" (Font & Mendez, 2013, p.15).…”