Research in this field covers a wide range of topics, including survey sampling, questionnaire design, privacy protection, and interviewer training. Stockholm University has a long tradition of research in this area, and several key publications in the field have been written by our researchers.
The field includes data collection methods, assessment of measurement errors, nonresponse issues, sampling, estimation including temporal issues of survey inference, register-based statistics, statistical disclosure control, response burden and total survey quality.
Central decisions by the Swedish government and our perceptions of society are based on official statistics (e.g is unemployment high or low); hence quality of the statistics is crucial and research into assessment and improvement of quality is a vibrant area.
Rising nonresponse levels, costs of traditional surveys and response burden issues makes efficiency particularly important. Random sampling has been gold standard since the early 20th century; however, with recent developments, current research focuses on rigorous inference based on nonrandom sampling, samples subject to missing data, register-based statistics, total survey quality and alternative data sources such as “big data”.
Read more about the history of official statistics in Sweden