Do open-ended survey questions on migration motives create coder variability problems?

Thomas Niedomysl, Bo Malmberg

Research output: Contribution to journalArticlepeer-review

27 Citations (SciVal)


Contemporary research on migration has benefi ted from adopting a variety of methodological approaches and different sources of information to provide answers to the ever-recurring question of why people migrate. Yet, when it comes to central methods used for researching migration motives, progress appears to have been slow. This paper focuses on surveys to research migration motives using self-administered postal questionnaires. It addresses a key validity
question, namely the issue of whether the usage of open-ended questions creates coder variability problems. An experimental
research design was used where fi ve coders independently coded 500 randomly selected responses from a large survey on migration
motives. Krippendorff’s a was calculated to test the level of agreement between the coders. The results advance our knowledge in two
important ways: fi rstly, it is shown that coder variability is not a major problem (Krippendorff’s a = 0.82). Secondly, it identifies those types of responses that nevertheless appear problematic to code. The implications of these fi ndings for survey research on migration motives are discussed, and it is argued that open-ended questions have some distinct advantages compared with the more commonly used closed-ended questions.
Original languageEnglish
Pages (from-to)79-87
JournalPopulation, Space and Place
Issue number1
Publication statusPublished - 2009

Subject classification (UKÄ)

  • Human Geography


  • migration motives
  • survey design
  • postal questionnaires
  • content analysis


Dive into the research topics of 'Do open-ended survey questions on migration motives create coder variability problems?'. Together they form a unique fingerprint.

Cite this