Statistical processing
Contact info
Short Term Statistics, Business StatisticsKaren Keller
+45 21 19 85 61
Get as PDF
The survey is made on the basis of a randomly selected sample with approx. 6000 interviews in net value (unit nonresponse excluded) among Danish residents aged 15 years+. The sample is then grossed up to the target population (Danish residents aged 15 years+) by using different strata, which refers to groupings based on for instance gender, age, family type, family income and socio group in order to make the sample represent the target population the best way possible.
Source data
Sample survey among Danish residents aged 15 years and older. Approx. 6000 interviews per annum.
Frequency of data collection
Data is collected on a daily basis.
Data collection
Mixed data collection method. Data are collected via telephone interviews (CATI) or web surveys (CAWI). The telephone interviews were previously conducted in-house by DST Survey, but since June 2019 they have been carried out by external survey providers, first Norstat and, since 2026, Wilke. Interviews are conducted on a daily basis with selected respondents. The web survey was introduced in 2013 and has become the predominant mode of data collection, gradually replacing telephone interviews. A new questionnaire was introduced in August 2017 and revised in June 2019 in connection with the transition to Norstat. A new questionnaire were introduced in 2025. In 2026, Wilke took over responsibility for the data collection from Norstat. Telephone interviews are still conducted, but since 2026 only among respondents aged 65 years and over. The primary mode of data collection is the web survey.
Data validation
Data is validated by comparing with previous reference periods as well as crosschecked for consistency. If errors or anomalies are detected they are examined and will either be discarded or imputed before aggregating the data.
Data compilation
Statistics Denmark collects data by interviewing a random sample of the Danish population aged 15 years and over. The anonymised interview data are validated and checked for errors. Among other things, consistency between variables is examined, and any outliers are removed. The respondent data are then combined into quarterly datasets, which form the basis for calculating calibration weights. The weights used to gross up the sample to the total population are calculated using the CLAN procedure, a standardised calibration method. Once the weights have been calculated, the four quarterly datasets are combined, and the sample is grossed up to represent the full population. The aggregated dataset is then validated again by comparing the results with previous periods and by checking consistency across variables.
Adjustment
Data is not adjusted after publication.