PhD scholarships

1. Learning analytics methods for classroom teaching in science education

1.1 Modelling of terminology acquisition in physics teaching

In the STEM fields, the writing of test protocols is one of the terminologically challenging tasks, that progressively should be improved from the 5th to the 13th grade in school. The use of machine evaluation methods and the modelling of learning progression can support teachers in correction and individual support. For this purpose, the subject-specific terminology is of particular importance. In the project, an application-oriented evaluation of protocols for physics will be achieved by building up a school-related terminology collection and by using methods such as artificial neural networks. We aim to transfer of the knowledge gained from physics to other STEM fields.


1.2 Ontology-based modelling of learner profiles and curricula

Open Educational Resources have the potential to significantly simplify the design of teaching materials. Therefore, however, it must be possible for teachers to find suitable resources efficiently. This project explores techniques how Semantic Web technologies and automatic information extraction methods may help to match a teacher’s current intent to  appropriate existing materials.


1.3 Semi-automatic classification of student drawings and texts in Science Education

Students’ individual conceptions about scientific processes can be diagnosed based on free-text descriptions and drawings produced during classes. However, their detailed analysis is a time-consuming task for teacher. This PhD project explores automatic methods of image and text analysis to support this working step, building on the integration of state-of-the-art machine learning techniques and semantic knowledge graphs.


2. Data-based support of students’ learning potential

2.1 Analysis of individual task processing (text and drawing) to prepare a collaborative work phase for developing conceptual understanding

This PhD project examines conditions of success of an automated analysis of different task formats (multiple select, open text and self-generated drawings) for an adequate assignment into small groups in a collaborative learning setting.


2.2 Data-analytics-based recommendations for individual learning

Modelling of a concept structure for a biological and/or chemical topic, which includes scientific and everyday understanding. Based on this, learning analytics methods are adapted, which allow a diagnosis of individual understandings and learning recommendations.


2.3 Formative assessment in the classroom supported by machine learning

In school lessons, large amounts of data are continuously accumulated, even with formative assessment methods, which can hardly be adequately evaluated and used by a teacher in a reasonable amount of time. In this project we investigate how such data can be efficiently evaluated using machine learning methods and how teachers and learners can use them in the teaching-learning process.


2.4 Promotion of the transition from instructed to independent problem solving in physics by learning resources automatically adapted to learning progress

Methods of Learning Analytics should be used in a targeted manner to provide secondary school students with learning materials that are individually adapted to their learning progress, for example to avoid the illusions of understanding typical for beginners or to promote the anticipatory approach typical for experts.


2.5 Using Learning Analytics for Investigating Learning Processes in Inclusive Chemistry Classrooms using Universal Design for Learning

A main argument for the use of digital media in science education is to enable more individualised learning. This research project focuses on how the use of iPads, for example, can be automatically analysed and supported in the classroom using learning analytics.


3. Transition from school to university

3.1 Development and evaluation of digital resources for the appropriation of structural praxeologies in the transition from school to university mathematics

For mathematics at the university, so-called structural concepts or praxeologies (Hausberger, 2018) are particularly important. Mathematical practices linked to these concepts and praxeologies concern, among other things, justifications with reference to object sets, relationships between object sets and structure-invariant operations. With regard to such praxeologies in analysis and higher mathematics for engineers, concrete proposals for digital support measures in the transition from school to university and the first year of study will be developed and evaluated and optimized in a Design-Based-Research approach.


3.2 Evaluation of the influence of learning environments on the effectiveness of bridging courses and the development of digital methods to support

The mathematical bridging courses at the Ostfalia University aim in the refreshing of basic mathematics. In addition the understanding of mathematical concepts and the learning behaviour should be improved. Therefore a special learning environment is used, where the students take the active part in the course. The effectiveness of this learning environment should be evaluated. Based on this digital procedures should be developed, which support the positive effects of the used methods.


4. Informal learning in schools and universities

4.1 Data Analytics for informal learning in school settings

Nowadays, many students use informal learning resources on the Web to explore new learning content. This PhD project records and analyses their search behaiour in controlled study settings. The objective is to identify behaviours beneficial to learning and to determine approaches to support them, in particular through individual recommendations of contents and study strategies.


4.2 Use of informal media for basic mathematics by university students

Development of learning environments at the university where informal media (like YouTube, etc.) are used. The aim is to increase the media literacy of the students and by this to improve the ability of learning mathematics. 


4.3 Analysis and modelling of the fostering of mathematical enculturation in the transition from school to university through processes of informal learning

Mathematical enculturation arises from newer, socio-cultural theories (Pepin 2014). It addresses the subject-related participation in so-called “authentic” activities of the new, university mathematical culture. This includes the adoption of beliefs, values, goals, procedures and an adaptation of one’s own identity. The contribution of digital informal learning activities to mathematical enculturation has not yet been investigated, which is also due to a lack of data. Methods of Learning Analytics offer new possibilities for the analysis and modelling of relevant processes.


5. Data protection and fairness of learning analytics methods

5.1 Dealing with bias and discrimination in learning analytics models

Data-driven decision making entails the risk of discriminaton towards specific individuals or population segments defined upon protected attributes like gender, race etc. In this project, we will investigate the sources of bias as well as methods for bias detection and mitigation of biased decision-making in the learning analytics domain.


5.2 User-friendly configuration of access control mechanisms for Learning Analytics

The PhD project focuses on the development and evaluation of procedures for the secure and user-friendly configuration of access control mechanisms for learning analytics data. The objective is to allow teachers without high levels of expertise in IT security and data protection to use student data effectively, while complying with data protection requirements.



  1. Hausberger, T. (2018). Structuralist Praxeologies as a Research Program on the Teaching and Learning of Abstract Algebra. International Journal of Research in Undergraduate Mathematics Education, 4(1), 74-93.
  2. Iosifidis, V. & Ntoutsi, E. (2018). Dealing with Bias via Data Augmentation in Supervised Learning Scenarios. In J. Bates, P. Clough, R. Jäschke, J. Otterbacher (Eds.) Proceedings of  the Workshop on Bias in Information, Algorithms, and Systems, in conjunction with iConference 2018, Sheffield, United Kingdom, CEUR, 24-29

%d bloggers like this: