A feasibility study for a survey of migrants

Publisher:
Home Office
Publication Type:
Report
Citation:
2011, pp. 1 - 46
Issue Date:
2011-01
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail2013000469OK.pdf Published version552.26 kB
Adobe PDF
Overview. The Analysis, Reserach and knowledge management Directorate (ARK) in the UK Border Agency commissioned Ipsos MORI and the Institute of Education to undertake a feasibility study for a large-scale face-to-face survey of migrants in the UK. The purpose of the feasibility study was to inform the design of a survey of migrants, looking at how to design and interview statistically robust sample and the appropriate questions to adopt such a survey. The feasibility study involved: 1. workshops with the UK Border Agency and other stakeholders to identify survey requirements; 2. the development of a definition of 'migrant' for use in the survey 3. the development of a sample design; 4. testing fieldwork recruitment and data collection methods to be used; and 5. the development of a questionnaire, including cognitive testing in several languages This report focuses on the sampling and field methods aspects of the feasibility study. A separate technical report provides details on other issues, including question development. To inform the development of the survey methodology and the survey questionnaire discussion groups and depth interviews were undertaken with representatives of economic migrant communities and individuals working with refugees and asylum seekers in England and Scotland. Key issues surrounding survey objectives, terminology, and field methods were identified. A useable survey population definition was developed after stakeholder consultaion such that, for the purpose of this study: 'A migrant is defined as someone who arrived in the UK in 1990 or later, who was a non-UK national on entry, whose usual place of residence prior to entry was not in the UK, and who has lived in the country for at least three months (one month if an asylum seeker or refugee).' It was concluded that different types of migrants will require different sampling and screening methods. Asylum sekers could be sampled directly from administrative records but other migrant types would have to be screened in the field. Small area estimation methods were successfully used to identify areas in which migrants are more heavily concentrated in order to inform the sample design. A number of field screening methods were investigated and field tested. Despite testing some innovative methods, traditional screening methods were found to still have an important role to play. It was concluded that non-asylum-seeker migrants should be screened through a combination of traditional screening methods for areas of higher migrant concentration and an adapted form of the Waksberg method (a two-stage method originally developed to minimise the number of calls made to ineligible numbers in telephone surveys) for areas of lower migrant concentration. An overall approach to a sample design was outlined and illustrative designs were presented applicable potentially to 6,000 interviews in England and 3,000 in Scotland. It was concluded that the survey should aim for 80 per cent coverage of the migrant population and that fieldwork effort should be disproportionately concentrated in areas of higher migrant concentration. The overall conclusion from the work was that a high quality survey of migrants would be feasible although a detailed design would need to take into consideration the level of funding available and how competing design requiremenets should be prioritised. In carrying out any such survey a large-scale dress rehearsal survey should be conducted before resources are fully committed in order to confirm the assumptions underlying the sample design and to field test the questionnaire and field translation/interpreting procedures.
Please use this identifier to cite or link to this item: