Accuracy and reliability
Contact infoResearch, Technology and Culture
+45 39 17 30 26
Get as PDF
The cultural habits survey is based on approx. 3,000 completed interviews per. quarter, which is a relatively robust sample that helps reduce the uncertainty of results. The uncertainty is further reduced by using a mixed data collection method and repeated inquiries to the respondents. The results of the survey can be compared with other statistics in several areas. These comparisons show a good coincidence.
An important source of uncertainty in this study is the sample uncertainty. Measured by key indicators, this source of uncertainty is less than one percent. There are answers from respectively 55 and 48 percent of those surveyed in the first two quarters of the survey, which is a further source of uncertainty for the study. It is also the case that certain types of groups more often do not respond and it affects the representativeness of the study. However, this is largely addressed through the enumeration and use of register-based assistance information. Although the assistance information tackles a lot of bias, it cannot be ruled out that there is systematic bias in the cultural consumption study.
Other sources of uncertainty, for example memory errors, are also relevant to the cultural consumption study. In order to reduce memory errors, respondents are questioned in recent weeks or in the past three months. This reduces the risks of memory errors in comparison with previous studies, which measured cultural consumption over a longer period, e.g. in the last year.
The precision of the cultural consumption survey can also be assessed by comparing selected results from the survey with relevant data from other statistics.
For example, the study estimates that six out of ten people stream music. The same proportion is calculated in the study 'IT use in the population 2018', where it is estimated that 63 per cent of the Danes between the ages of 16 and 89 stream music. The cultural consumption survey also shows that approx. 52 per cent of the Danes see movies or series via paid streaming services e.g. Netflix. The proportion is estimated at 50 per cent in the statistics "IT use in the population 2018".
27 per cent of the Danes state in the cultural consumption study that they use running watch, pedometer etc. when exercising. This proportion can be compared with the latest version of the statistics "Electronics in the home 2018". Here, it is estimated that 30 per cent of families own activity clock. Finally, the statistics on electronics in the home show that 9 per cent of the families are in possession of e-book readers. A share that is in line with the Cultural consumption survey's estimation for the proportion of Danes reading literature via e-book reader (10 per cent).
Based on the above-mentioned comparisons, the accuracy of the Cultural consumptionstudy seems to be high for the relevant indicators.
Sample uncertainty is measured by standard error, which expresses the uncertainty on the average of the estimates. The sample uncertainty is affected by the size of the sample. The uncertainty is reduced by half, for example, substantially if you quadruple your interview basis. The more responses the survey results are based on, the more secure the numbers are.
The standard error for selected indicators is less than one percent in the first two data collection rounds of the Cultural consumption survey.
In order to cover the population as best as possible, the sample for the survey is drawn very close to the start time of the data collection. It is a comprehensive questionnaire the respondents are faced with. This can lead to unwillingness to answer and thus uncertainty in relation to the extent of cultural consumption reported. It can be difficult for the respondents to remember what they have done, for example, three months back in time, which can also give uncertainty in relation to the extent of cultural consumption that is reported. Some forms of cultural consumption are perceived by many as "better" than others - for example, it is often perceived as better to read books or go to a museum than to watch movies or series on Netflix. This can lead to certain respondents wishing to give a better impression of themselves, which is also a source of uncertainty. The extent of these sources of uncertainty is difficult to measure. As the questionnaire does not contain particularly sensitive questions or questions of a complex technical nature, it is assumed that the extent of this type of error is limited.
An additional source of security is outdated information about the respondents' background information, eg education, family type or socio-economic status. Background variables are therefore drawn from Statistics Denmark's various statistical registers as close to the survey reference period as possible. However, there is always a subset of the sample where the register information is no longer accurate at the time of data collection, eg due to recent changes in education level or family composition. Dropout in the study can also be a source of uncertainty. Several measures have been taken to reduce the dropout rate in the study. Thus, there is collection of information both via a web form and via telephone interviews. Follow-up via interviews is especially used on groups that traditionally have greater dropouts. COVID-19 is not expected to affect the statistical sources of uncertainty. The data collection in 2020-21 proceeded as it used to. 2020 and 2021 data were collected during the closure of Denmark - which also affected participation in a number of cultural activities. This explains why the figures may vary from previous quarters. Annual data 2019 covers the period 3rd quarter 2019 - 2nd quarter 2020. 2nd quarter 2020-2021 was collected during the closure of Denmark due to COVID-19. This may explain why the figures for 2019 and 2020 may vary from the previous annual data set (2018).
Statistics Denmark follows the recommendations on organisation and management of quality given in the Code of Practice for European Statistics (CoP) and the implementation guidelines given in the Quality Assurance Framework of the European Statistical System (QAF). A Working Group on Quality and a central quality assurance function have been established to continuously carry through control of products and processes.
Statistics Denmark follows the principles in the Code of Practice for European Statistics (CoP) and uses the Quality Assurance Framework of the European Statistical System (QAF) for the implementation of the principles. This involves continuous decentralized and central control of products and processes based on documentation following international standards. The central quality assurance function reports to the Working Group on Quality. Reports include suggestions for improvement that are assessed, decided and subsequently implemented.
The study is considered relevant as the questionnaire has been developed in collaboration with users. Reliability is ensured by having openness about metadata, such as response rate, dropout, uncertainty calculations and any revisions and detailed information on processing data. The statistics can only be compared to a lesser extent with previous cultural consumption studies. This is due, firstly, to a method change where the reference period has changed from one year to the last three months. Secondly, the questionnaire is revised in relation to previous questionnaires to more accurately reflect today's cultural consumption, which is often digital. Thus, the statistics are more up to date and published more frequently (every quarter). Finally, it should be noted that in the period 2018-2020 minor adjustments are expected in the questionnaire in order to ensure a more optimal coverage of the target population's cultural and media consumption as well as leisure activities.
Data revision - policy
Statistics Denmark revises published figures in accordance with the Revision Policy for Statistics Denmark. The common procedures and principles of the Revision Policy are for some statistics supplemented by a specific revision practice.
Data revision practice
There are only calculated final figures.