Significance of the Institutional Context for Drop-Out and Long-Term Studies

Research question/goal: 

Significance of the Institutional Context for Drop-Out and Long-Term Studies is a collaborative project funded by the Federal Ministry of Education and Research (BMBF) in the research field of "prevention and intervention measures in higher education to reduce drop-out". Cooperation partners are the German Centre for Higher Education Research and Science Studies (DZHW), whose scientific director Prof. Monika Jungbauer-Gans coordinates the project, and the University of Hanover, represented by the project directors Prof. Christoph Hönnige (Political Science) and Prof. Volker Epping (Law; President of the University of Hanover). The subproject Institutional Context at the State and University Level was carried out at the MZES between 2021 and 2022. 

The joint project investigates the effect of the study-related institutional context on the course of study, intention to drop out, long-term study, and drop-out. The substudies Institutional Context, Duration of Studies and Drop-Out, and Intention to Drop Out examine the aforementioned effects on the basis of data from official statistics and surveys. The characteristics of the institutional context at three levels (state, university, degree programme) are determined by the partners at the University of Hanover and the MZES. Our substudy has used quantitative text analysis to measure the contextual feature space on a common theoretical basis.

To this end, the Mannheim subproject team has developed an elaborate coding scheme to capture the level of flexibility in examination regulations. The manual coding is used to train machine learning algorithms (especially Bidirectional Encoder Representations from Transformers, short BERT) to measure the character of other texts. A Shiny Server application has been set up on a virtual machine at the MZES to make the corpus, the coding, and the results of the machine learning available (see Two publications on the data or “BiK-Korpus” and the machine learning methods used are in preparation as direct outcomes of the Mannheim subproject. The director of the subproject based at the MZES will continue to be associated with the collaborative project until its conclusion.

Fact sheet

2021 to 2022
Data Sources: 
Laws and regulations at the state and university level
Geographic Space: 
German Bundesländer