German students abroad
Males and females of German nationality enrolled at a foreign higher education institution in any relevant year under report.
The Higher Education Statistics Act states that such students must be recorded. However, this ruling only enables those to be registered who were (also) enrolled at a German higher education institution during the survey period. This is quite frequently the case in degree programmes which involve agreed cooperation between a German and a foreign higher education institution. However, higher education institutions apply differing administrative procedures here, which means that students studying abroad are only incompletely registered by the official higher education statistics. The results at hand do not correspond with the real situation. Moreover, the official student statistics are unable to record Germans studying abroad who have not yet or have never studied at a German higher education institution or who - partly deviating from the data collection guide-lines defined by the official statistics - were dematriculated for the duration of their studies abroad.
This is why the Federal Statistical Office annually carries out a so-called Country Survey on behalf of the BMBF. This survey requests information from the statistical offices responsible for collecting education statistics in countries in which more than 125 Germans are studying in the period under report. Wissenschaft weltoffen presents the results of this survey. Some estimates have to be made for this survey, while some individual states do not (yet) have data available for the period under report. Moreover, problems may be encountered in how a country defines facilities and establishments that are part of the higher education sector and how it defines students. For example, the official statistics of some few countries (e.g. Italy) do not record German students who are attending a higher education institution in Italy under the ERASMUS programme as students.
The results of the Country Survey are nevertheless so close to realistic approximate values that they can be used.