Statistical processing
Contact info
Labour and Income, Social Statistics.Martin Faris Sawaed Nielsen
+45 23 69 90 67
Get as PDF
The subjective data from the interviews are combined with the register based data using the Central personal Register. To adjust for non-response bias, weights are computed and assigned to respondents. This ensures that the survey population match the Danish population on demographics and income levels.
Source data
The EU-SILC is a survey which combines interviews with administrative data from the registers at Statistics Denmark.
The primary source for SILC data is interviews with 6,200 households from a sample of 12.100 households. This sample is a combination of persons who participated the previous years and 7.000 added households. The interview data are combined with administrative registers to form the EU-SILC datasets. Denmark uses a selected respondent model and only interview one person per household. The selected respondent is asked personal questions, question related to the household and labor market status for all household members.
From 2016 the sample is stratified on Regions. From 2020 the sample is stratified on both Regions and the income intervals 0-60 per cent of the median, 60-100 per cent of the median and above 100 per cent of the median.
The target population is "persons living in Denmark", while the survey population is "persons living in private households in Denmark". Thus, persons living in institutions, prisons and the homeless are not included in the survey.
Frequency of data collection
Survey data is collected every year from February to May.
Data collection
80 per cent of the interview data is collected via web interviews, the rest is collected via telephone interviews. If the selected respondent have not responded within a shorter period a letter are mailed out reminding on the survey. If this does not give any reply the respondent are called by phone and the interview is conducted. Afterwards the data are supplemented by register based data.
Data validation
The interview data are validated at the interviewing where not existing codes cannot be used and warnings minimize the risk of incorrect data input. When asked to provide amounts, i.e. the monthly rent, filters give the respondent the option to change his answer, if the answer differs to much from the norm. Software with a range of data check provided by Eurostat ensures data are internally and externally consistency.
Data compilation
The respondents are assigned weights to correct for non-response bias. The weights are constructed using register data. The household respondent defines the composition of his/her own household. Assisted by an anonymous personal identifier the survey data are linked with registers that provides data on age, income, education etc. for all household members.
Adjustment
Not relevant for this statistics.