New Pub: Design, Development and Evaluation of HILA

New Pub: Design, Development and Evaluation of HILA

Artificial Intelligence, Keynote, Learning Analytics, Publication
How can AI-supported learning analytics be integrated into educational processes in a significant way?  How can they be designed, tested and further developed to effectively improve teaching and learning practices? These questions were addressed by Hendrik Drachsler in his keynote at the Learning AID 2024 in Bochum, which has recently been published in the conference proceeding “Learning Analytics, Artificial Intelligence und Data Mining in der Hochschulbildung”. In his keynote, Hendrik stresses the importance of content-specific applications that address genuine educational needs and are supported by empirical evidence demonstrating their effectiveness. The key to fostering adaptive and sustainable learning experiences is to understand and accommodate learners’ individual needs. Hendrik argues that technological progress alone is not sufficient to improve education. His ongoing research shows that AI-supported learning analytics can only bring…
Read More
New Pub: TBA at BEA 2025 Shared Task: Transfer-Learning from DARE-TIES Merged Models for the Pedagogical Ability Assessment of LLM-Powered Math Tutors

New Pub: TBA at BEA 2025 Shared Task: Transfer-Learning from DARE-TIES Merged Models for the Pedagogical Ability Assessment of LLM-Powered Math Tutors

New Pub
In the BEA 2025 Shared Task on Pedagogical Ability Assessment of AI-Powered Tutors, the goal was to evaluate how well LLM-based math tutors support students. The task focused on four aspects of feedback: spotting mistakes identifying where the mistake happens giving guidance providing actionable suggestions For our submission, we built on FLAN-T5 models with a multi-step training pipeline. In addition to standard fine-tuning, we used model merging (DARE-TIES) to leverage information across all four labels – and saw clear improvements over plain fine-tuning. Our models achieved F1 scores between 52 and 69 and accuracies between 62% and 87%, ranking 11th, 8th, 11th, and 9th across the four tracks. Link to the paper: Link
Read More
New Pub: How Feedback Literacy Moderates Student Perceptions of Feedback

New Pub: How Feedback Literacy Moderates Student Perceptions of Feedback

Feedback, Higher Education, Journal, New Pub
Whether or not students receive effective feedback can have a big impact on their overall learning process. The more qualitative and personalized the feedback is, the more effective it is in supporting the students with their individual learning goals. The latest developments in learning analytics and artificial intelligence have made it possible to provide a large number of students with personalized feedback automatically and simultaneously. Despite these technical advances, there is still much to be learned about the ability of students to use this feedback for their scholastic benefit. So far, only a limited number of studies have examined the impact of feedback literacy on students' perceptions of feedback, which is particularly true for technology-enhanced learning environments. A newly published paper addresses this issue and focuses on how students interpret…
Read More
New Pub: How Learners Interpret and Respond to Feedback in Learning Analytics Dashboards

New Pub: How Learners Interpret and Respond to Feedback in Learning Analytics Dashboards

Feedback, Journal, New Pub, Self-Regulation
Online learning platforms can generate a great amount of data about how students engage in the learning process. This data is used to develop learning analytics dashboards as feedback tools to assist students in self-regulating their learning. But how do students use these tools to self-reflect? And how can they use the information provided by learning analytics dashboards (LAD) in a meaningful way? In a new publication these questions are explored using the data from an experimental study with 417 students, which investigated how the students interpret and respond to feedback from LADs. In the study,  the students were divided into 2 groups: A treatment group was given personalized self-regulated learning (SRL) feedback from the LAD on the basis of interactions and progress; A control group was given minimal feedback…
Read More
New Pub: Tracking students’ progression in developing understanding of energy using AI technologies

New Pub: Tracking students’ progression in developing understanding of energy using AI technologies

Artificial Intelligence, Journal, Publication, School
[caption id="attachment_7553" align="alignright" width="400"] Instructional unit, with pre- and post-test, as well as lesson-set-level assessment[/caption] In physics education, some students fail to have the foundational knowledge of energy concepts needed to engage in societal debates on climate change and energy transformation. A newly published study highlights the potential of AI to identify students with different learning trajectories and to help bridge the knowledge gaps. The researchers used a digital workbook designed to teach energy concepts to collect detailed interaction data from over 500 students. After applying exclusion criteria, data from 172 students were analyzed to identify their productive and unproductive learning curves. [caption id="attachment_7554" align="alignleft" width="400"] Example single choice pretest item[/caption] By using machine learning, specifically random forest models, and natural language processing (NLP), the researchers were able to classify…
Read More
A Warm Welcome to Our Guest Researcher Yildiz Uzun

A Warm Welcome to Our Guest Researcher Yildiz Uzun

New Pub
We are delighted to welcome Yildiz Uzun as a guest researcher in our team! Yildiz is a PhD student at the Knowledge Lab, Institute of Education (IOE), University College London (UCL). Her research explores how students regulate their learning with the support of learning analytics feedback in the form of a dashboard, with a particular focus on how they engage with this feedback. She also examines how engagement with analytics feedback can be scaffolded and enhanced through the integration of an Artificial Intelligence (AI) assistant. During her visit to DIPF, she will collaborate with Prof. Hendrik Drachsler and the EduTec team to advance her current AI-driven feedback process by analysing student queries to the AI assistant, evaluating the pedagogical quality of AI responses and identifying ways to optimise conversational systems…
Read More
[Workshop] Introduction to Language Technology and Language Modeling for Education

[Workshop] Introduction to Language Technology and Language Modeling for Education

New Pub
At the recent 19th Joint Summer School on Technology-Enhanced Learning located in Rhetimno, Greece, I (Sebastian Gombert) gave an introductory workshop on language technology and language modeling and their various use cases of in education. This included use cases such as short answer scoring, essay scoring, classification of texts according to the CEFR framework, and group communication analysis in CSCL. Overall, the workshop was well-received and well-attended.
Read More
New Pub: Pedagogical Framework for Hybrid Intelligent Feedback

New Pub: Pedagogical Framework for Hybrid Intelligent Feedback

Artificial Intelligence, Feedback, Journal, Publication
[caption id="attachment_7525" align="alignright" width="300"] Principles of hybrid intelligent Feedback[/caption] ChatGPT and other Generative Artificial Intelligence (GenAI) tools have quickly become an active part of everyday academic life. Both teachers and students use them regularly to do research, summarize texts or responses, create exercises or quizzes or explain course content. But GenAI needs to be used in a sensible manner and consistent with sound teaching principles. Otherwise, it could lead to an over-reliance on technology and neglect the important roles of teachers and students. Ideally, GenAI should be used to complement rather than replace human-centered education. When pedagogical practices combine the strengths of humans and AI, this can ultimately lead to enhanced creativity, empathy and academic comprehension of students. A recently published paper explores a promising field of human-AI collaboration in…
Read More
New Pub: On-Your Marks, Ready? Exploring the User Experience of a VR Application for Runners

New Pub: On-Your Marks, Ready? Exploring the User Experience of a VR Application for Runners

Multimodal Learning Analytics, Publication
Do you wish to be a better runner and ready to get going? Becoming a better runner is more than just choosing the right pair of running shoes and increasing your physical ability to perform a successful run. You also need to be mentally prepared to increase your running productivity. Here is where Virtual Reality (VR) and cognitive behavioral therapy (CBT) can help. Up until now, there have been limited options available for mental training using technologies like sensors and actuators. These technologies have primarily been used to improve physical training. A newly published study explores the possibilities that VR and CBT have to create a technology-based mental training for athletes. Our team carried out a formative study to explore the user experience by incorporating CBT techniques into VR technology,…
Read More
New Pub: Enhancing User Onboarding in Virtual Reality Educational Applications

New Pub: Enhancing User Onboarding in Virtual Reality Educational Applications

Multimodal Learning Analytics, Publication
Learning in a Virtual Reality (VR) environment can be a captivating and immersive experience for students. Despite the potential that VR has to help students learn in new and exciting ways, many students may also struggle during their initial interactions with a given VR environment. Whether or not effective user onboarding takes place can be the make or break of an application’s success. A new publication sheds light on this key challenge of effective onboarding by exploring how pre-training methods impact the onboarding experience in VR educational applications. 36 people, primarily university students, participated in a study to understand how different approaches affected their ability to navigate VR environments and their overall satisfaction. The findings show that pre-training helps users get an orientation and slightly boosts satisfaction, but it falls…
Read More