The data for these statistics are collected yearly in October from the administrative systems of the educational institutions. Statistics Denmark have drawn up standards for reporting formats, validation and error searching in cooperation with the system provider SIS, who is in charge of the administrative systems at most institutions who offer medium-cycle higher education. This means that the data received by Statistics Denmark have already been controlled. In addition, a thorough error search and validation of data are done by Statistics Denmark, where data is controlled for fluctuations across time and irregularities at the individual level.
The sources for the statistics are data from the educational institutions' administrative systems.
The statistics are based on information collected once a year.
The majority of data is collected by system to-system reporting through the administrative system SIS. Some institutions report manually through spreadsheets.
The data validation mainly consists of two types of error searching: macro error searching and micro error searching.
Macro error searching consists in searching after fluctuations on four parameters: Population (the number of students per 1 October), Entrants, Completions, and Discontinuations. The numbers are compared with the numbers from last year for every education on the institution. If bigger and relatively unlikely fluctuations are found, the institution have to either correct the numbers or tell Statistics Denmark that the numbers are correct and shows the factual development.
Micro error searching focuses on errors in the data received by Statistics Denmark about specific persons. Examples of micro errors could be an invalid education code or the student's start date being after the end date. A check for overlaps are also made, since there is not always agreement between the dates provided when the students change institutions. Statistics Denmark corrects the dates so the student does not seem to be enrolled on two institutions at the same time.
Data from the institutions are aggregated to a total data set. Data are not processed much by Statistics Denmark, since the quality of the data is very good.
No corrections of data are made except what has already been described under Datavalidering and Databehandling.