Contact infoResearch, Technology and Culture
+45 39 17 30 26
Get as PDF
The cultural consumption survey is a sample-based questionnaire survey. Approx. 1,000 responses are collected each month either via web form or as a telephone interview. The results are corrected for conditions, such as dropouts and the composition of the population. Data quality is checked at both micro and macro level. During the data processing, a limited number of responses are corrected due to non-consistent or obviously incorrect answers. The year 2018 covers 3rd quarter 2018- 2nd quarter 2019. The year 2019 covers 3rd quarter 2019-2nd quarter 2020. The year 2020 covers 3rd quarter 2020-2nd quarter 2021.
The statistics are prepared on the basis of sample-based questionnaires. The sample is drawn quarterly from The Central Person Register (CPR), which is updated daily. Approx. 1,000 completed responses are collected via telephone interviews or web questionnaire each month. The frame population is the same for each quarters covering persons 16 years old or older with residence in Denmark.
Frequency of data collection
The data collection is carried out every month throughout the year combining computer-aided telephone interviews (CATI) and web-gathering. The collection method is distributed at 28 per cent CATI and 72 per cent web in the first round of collection in 3 quarter 2018. The corresponding distribution for 4 quarter 2018 is 31 per cent CATI interviews and 69 per cent Web reports. In total, approx. 2,000 people each month. The goal is approx. 1,000 completed interviews per month and at least 10,000 completed interviews per year.
Data validation takes place partly during the data collection as non-consistent responses are reduced by using routing. This means that the interviewee is only asked about e.g. number of concert visits, if he or she stated in previous questions that they have used "concerts" as a way to listen to music in the past three months. Furthermore, data is validated after data collection using consistency checks and comparison with previous data collection results. Finally, the so-called macro error search focuses on the items that have the greatest impact on the finished statistics product. Key variables are compared to results of other relevant studies.
Calculating weights and enumerating the entire population of results is an essential part of data processing. The weights are used to enumerate the results of the sample survey to make them representative of the entire target population and to adjust for discontinuations. The weights are calculated using the assistance information on the interviewed persons from statistics registers using regression models. Grouped register variables such as gender, age, education, family income, socioeconomic status and residence address, grouped as degree of urbanization, are part of the enumeration model.
Data is also processed in the form of the creation of derived variables. For example, derivative variables, that indicate whether the individual cultural activities have been "used" over the past three months, are calculated. The number of rhythmic concerts is calculated by subtracting the number of classical concerts (Music6) from the number of all kinds of concerts (Music5).
Not relevant for these statistics.