Accuracy and reliability

Overall accuracy

In general, the quality of the statistics is rated highly.

For the reference year 2021 there are 29 enterprises in the sample, covering almost 100 per cent of the total turnover in the sector for Market research and public opinion polling. 28 enterprises answered the questionnaire, i.e. the response rate is 97 percent. The usable questionnaires are from 28 enterprises. There is a lapse of 3 per cent. The lapse are due to wrong industry code and closure of the enterprises and non-responded questionnaires.

The usable data from 28 enterprises grossed up to the total level of the 6-digit industries within the sector for Market research and public opinion polling by using preliminary turnover from the Business Register. Each company is assigned a weight, that corresponds to the ratio of population and responses in a given size group and industry. The enumeration is made by ratio estimation.

As the survey is random, the figures can be associated with some statistical uncertainty in the form of random variation. The sampling error is calculated for all services in the form of variance, 95 pct. confidence interval and coefficient of variation Uncertainty estimates can be seen in the attached document usikkerhedsestimater_MENI.

The figures for 2021 are preliminary and will not revised in the national publishing. For delivery to EU the figures are calculated (revised) based on Accounts Statistics.

Sampling error

The sample was drawn at Jun 2021 consisting of 29 enterprises. Of these, 28 companies have completed the questionnaire, i.e. the response rate is 97 percent. The usable questionnaires were from 28 enterprises.

The sampling error is calculated for all services in the form of variance, 95% confidence interval and coefficient of variation Uncertainty estimates can be seen in the attached document usikkerhedsestimater_MENI.

Non-sampling error

The total uncertainty consists of sampling uncertainty and other systematic uncertainty. In other uncertainties include inter alia measurement error and non response error/lapse.

Coverage error

The population is defined by the industry, the enterprises registered in the Business Register by themselves. Errors of classification of the industries will be expected. This means that there may be enterprises in the population that really should not be classified as an enterprise in the sector Market research and public opinion polling.

In the context of fault tracing and quality assurance of the statistics, number of these errors continuously are found and corrected. Companies that incorrectly classified as Market research and public opinion polling cause over-coverage. These coverage error is detected primarily for companies in the sample, but also in some cases for non-sampled companies.

Under- coverage in the form of companies that are classified entirely outside Market research and public opinion polling, even though they should properly be classified as Market research and public opinion polling, are difficult to detect, but can be found also.

The total number of enterprises in the population is estimated at 29, and among these the sample is selected. The sample for the reference year 2021 was 29 companies and is thus 100 per cent. of the total number of the population. Because of the stratified selection, where large companies are more likely to selection than the smaller, the sample consists 100 per cent. of the total turnover of the population.

Measurement error

Uncertainty due to lapse, incorrectly reported numbers and misunderstandings has sought to minimize by repeated reminders by non-reporting, as well as by substantial verification of the reported figures.

Generally the services in the questionnaire are based on Statistical Classification of Products by Activity CPA 2008, with an adjustment to Danish conditions. It can be an uncertainty in the distribution of turnover and export by services when the services CPA 2008 are not quite commensurate with the classification of products of the firms.

Non response error

There are 29 enterprises in the sample. The usable questionnaires are from 28 enterprises, i.e. there is a lapse of 3 per cent.

Quality management

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.

Quality assurance

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.

Quality assessment

In general, the quality of the statistics is rated highly, with a few reservations. It can be problems in comparing figures over time. The survey was conducted for the first time in 2001. The years 2005 and 2007 are not comparable because of a shift to NACE Rev. 2 activity classification. The years 2007 and 2009 are not directly comparable because of changes of in the products in the questionnaire which is adjusted to the EC-questionnaire.

From the reference year 2013 the method is changed so the enumeration is based on a new data from the Central Business Register. That means the results before the year 2013 are not directly comparable to the results from 2013 onwards.

For the reference year 2021 there are 29 enterprises in the sample, covering 100 per cent of the total turnover in the sector for Market research and public opinion polling. 28 enterprises answered the questionnaire, i.e. the response rate is 97 percent. The usable questionnaires are from 28 enterprises. There is a lapse of 3 per cent. The lapse are due to wrong industry code and closure of the enterprises and non-responded questionnaires.

The usable data from 28 enterprises grossed up to the total level of the 6-digit industries within the sector for Market research and public opinion polling by using preliminary turnover from the Business Register. Each company is assigned a weight, that corresponds to the ratio of population and responses in a given size group and industry. The enumeration is made by ratio estimation.

As the survey is random, the figures can be associated with some statistical uncertainty in the form of random variation. The sampling error is calculated for all services in the form of variance, 95 pct. confidence interval and coefficient of variation Uncertainty estimates can be seen in the attached document usikkerhedsestimater_MENI.

The figures for 2021 are preliminary and will not revised in the national publishing. For delivery to EU the figures are calculated (revised) based on Accounts Statistics.

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

The published data will not be revised, since the statistics is published once every second year.