Accuracy and reliability

Overall accuracy

The statistics are considered a reliable measure of the population, as municipalities themselves report and confirm the aggregated figures. In addition, Statistics Denmark conducts various checks for potential duplicates and other errors that may be overlooked by the municipalities. Upon publication, there are typically minor instances of non-response and measurement errors in reports from some municipalities. However, known errors are usually corrected in subsequent releases and follow-up validations. Overall, with each re-release of previous years' data, the error rate is typically below 1 pct.

Every municipality is obliged to evaluate and approve the total amount of notifications for each year to confirm the data quality before the release. It is worth noting that Læsø municipality reports their received notifications of concern via Frederikshavn municipality.

Status on municipal approvals: For 2023, 2022, 2021, 2020, and 2019, all 98 municipalities have approved the reporting figures. For 2018, 94 municipalities have approved the data. Aalborg, Gribskov, Horsens, and Sønderborg municipalities have not approved the figures. For both Gribskov and Sønderborg, the number of reports is underestimated. For 2017, 97 municipalities have approved the data, with Aalborg unable to approve due to IT-related issues.

Note for 2023:

Indicated causes for the notifications, that have been gradually phased out during 2022, are present to a lesser extent also in the 2023 data. This is due to IT challenges during the transition to new specification requirements in some municipalities. These expired causes are:

• No. 7: Abuse against a child or young person, e.g., sexual or violent
• No. 8: Other forms of neglect towards a child or young person
• No. 16: Other

These discrepancies are expected to be resolved in the next validation and subsequent release of the statistics.

Odense Municipality is underestimated by 90 reports (approximately 2 pct. of the total for the municipality) due to issues with the KMD-NEXUS file system. These issues are also expected to be corrected during the upcoming validation.

Tønder Municipality has reported an increase in the number of notifications in 2023 due to improvements in reporting practices related to reports involving foreign nationals residing in Denmark.

Comments regarding 2022: The amount of notifications for Næstved Municipality is slightly off-estimated, approx.. 60 cases.

Comments regarding 2021: The amount of notifications for Randers Municipality is slightly underestimated ca. 70 cases. The Municipality of Syddjurs has identified and removed ca. 300 duplicates. The total amount of notifications for Syddjurs in 2021 is therefore lesser compared to previous periods.

Comments regarding 2020: The number of notifications in Næstved Municipality is underestimated, ca. 137 cases.

Comments regarding 2019: In 2019, it was discovered that at least one municipality had a practice of registering one cause per notification - also in those instances where several causes were the case. The formulation regarding the correct practice of indicating the causes of a given notification has been clarified in the Executive Order on Transmission of Data in the Social Policy Area and communicated to the municipalities. It is therefore expected that this cause of underestimation will be minimal in the future.

In 2019, Statistics Denmark has been made aware that at least one municipality (Randers Municipality) registers notifications of concern in 2 different departments, resulting in one department not reporting their notifications of concern to Statistics Denmark during the whole period. The overall amount of notifications for Randers Municipality in 2019 is therefore underestimated by 39 notifications.

In 2019, Statistics Denmark has been made aware that notifications of concern regarding Læsø Municipality in the period 2016-2018, were calculated as notifications from Frederikshavn Municipality. The number of these is low, and the error is/will be corrected in the future.

Comments regarding 2018: The Municipalities of Gribskov, Horsens, and Sønderborg, have not approved their data. In both Gribskov and Sønderborg's cases, the total number of notifications is underestimated. In the case of Horsens Municipality, the number can be over- or underestimated.

Comments regarding 2017: The amount of notifications for Ikast-Brande Municipality is overestimated.

Sampling error

The sampling error is zero since the statistic's method is complete enumeration.

Non-sampling error

Before 2017, inter-municipal notifications contributed to some duplicates.

The relatively large number of notifications can contribute to a minor risk of overseeing registration errors of individual cases. The validation process can catch many mistakes in total, but minor problems can occur in the totals between categories.

The large number of notification causes can contribute to an underestimation of the total number of causes. Furthermore, differing registration practices, in which the municipalities interpret the causes differently, can weaken the possibility to compare the causes of notifications.

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

The statistics are accurate and reliable since it is a complete enumeration, and due to the high quality of data. The data quality is ensured through the extensive validation process and annual revisions of earlier published data. The ongoing process of ensuring the quality of data can contribute to minor fluctuations that are under 1 pct. in total. Such fluctuations can occur, for instance, if errors, or non-approved data, stemming from previous years, are corrected upon a current publication of the statistics. These instances occur especially if a municipality has not managed to approve their data upon a given publication and corrects these issues upon the subsequent publication of the statistics. On an overall level, such fluctuations are minimal, and the statistics are thus considered accurate and reliable.

The gradual process of improvement in data quality has contributed to the fact that comparisons between the earliest and latest periods of the statistics are less accurate than, for instance, comparisons between two years in a row. The improved data quality is a result of better practices and more precise measures of errors in the validation process. However, the data, on an overall level, are comparable over time, when taking into account, the reorganization process completed in 2017.

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 statistics are revised annually two years back in time. New numbers are published at the same time as the previous two years and are republished with corrections, which have been reported since the last publication. For example, in 2024, data for the census year 2023 and revised data for 2022 and 2021 are published. Republication of previous years consists mainly of minor corrections and does not change the overall picture. Data is generally considered finite 3 years after the publication.