Publication Archives - EduTec Science

Nov22026 by Zhifan SunNo Comments

[NEW PUB] Two EduTech papers presented at the main track of EMNLP Conference

Two papers from EduTech have been accepted to the main track of the 2026 Empirical Methods in Natural Language Conference (EMNLP) and were presented at the conference in Budapest, Hungary. Paper1: Alice: A Large-Scale German Benchmark for Rubric-Based Multi-Dimensional Automatic Short Answer Scoring This paper presents the full Alice Dataset, a large-scale German student answer scoring benchmark. The novelty of the benchmark lies in the multidimensional scoring, where each student answer is scored according to three subtasks: Learning Performance, which measures how well the student answers the question directly; Knowledge Elements, where each answer is scored by how well the student masters and utilises the target concepts for each question; and Skill, which is the cognitive concepts required by each question. In addition, each available level in three subtasks is…

Oct282026 by Sebastian GombertNo Comments

[New Pub] How Do I Say It Best? What Prompt Optimisation Can Reveal About Optimal Prompt Formulation In LLM-based Automated Short Answer Scoring

New Pub

How much does the wording of a prompt matter when AI scores students’ answers? In our paper, we investigated this using three open-weight language models and two English-language short-answer scoring benchmarks. We compared human-written prompts with prompts automatically refined using ProTeGi, which analyses scoring errors and proposes improved instructions. Across 135 paired comparisons, optimisation delivered a median improvement of around two F1 points. Beyond improving performance, it also revealed patterns in effective prompt design: optimised prompts used more words that make logical relationships explicit, such as “if” and “because,” and shorter noun phrases. These features were also associated with better scoring performance, while more directive verbs were associated with poorer results. Our findings suggest that clear logical structure and simpler phrasing are promising principles for prompt design, and that automatic…

Oct192026 by Heather WehnerNo Comments

New Pub: How Students Think About and Act Upon Personalized, Automated Feedback on Transactivity

Computer-supported collaborative learning, Conference, Feedback, Higher Education, Publication

A key factor for success in collaborative learning environments is transactivity. This process, whereby learners engage with and build upon the ideas of others, is crucial for learning outcomes. In today's digital learning environments, AI-powered speech and text analysis can be used to measure the extent to which learners interact with one another. In a recent randomized field study involving 252 student teachers, a research team examined how different forms of automated feedback influence the transactivity in online discussions. The study compared basic feedback (BAF), consisting of general information and a performance summary, with highly informative feedback (HIF), consisting of highly personalized, detailed feedback on various aspects of transavtivity. The results were surprising: Although the participants rated the personalized feedback significantly more positively, it did not lead to any measurable…

Oct72026 by Prof. Dr. Hendrik DrachslerNo Comments

Keynote: 5 years of Highly Informative Learning Analytics

Artificial Intelligence, Conference, Feedback, Higher Education, Keynote, Learning Analytics, Learning Design, Research Methods, School

Five years in forty minutes On 7 October 2026, Hendrik Drachsler gave a keynote at the CBIE 2026 conference in Goiana, Brazil, that tried something unusual: time travel. He compressed five years of field research, seven projects and roughly 5,750 students into one forty-minute talk. The question behind the keynote was simple: after five years of developing and testing Highly Informative Learning Analytics (HILA) in real lecture halls, what do we actually know? Not what we hoped HILA might achieve back in 2020, but what five years of our own research and data now tell us. What HILA set out to do HILA starts with a modest observation: most lecture rooms still offer very little personalisation. Learning analytics promised to change that, but its rollout in Europe stalled on four…

Sep302026 by GianlucaNo Comments

mISPL: Mini Framework for Intelligent Systems for Psychomotor Learning @ ECTEL 26

Conference, Workshop

I recently had the pleasure of presenting my latest work at the MILeS 2026 workshop (Fourth International Workshop on Multimodal Immersive Learning Systems), held during the ECTEL 2026 conference in Valencia, Spain. It was an incredible week of pushing the boundaries of technology-enhanced learning, exchanging ideas, and building lasting professional bonds. Here is a quick recap of my experience, from early-stage frameworks to visionary futures and unforgettable social gatherings. Part 1: Introducing mISPL In the first part of the workshop, I presented my paper: "mISPL: Mini Framework for Intelligent Systems for Psychomotor Learning". While comprehensive frameworks like PsyLearn exist for building full-fledged psychomotor tutoring systems, they can often feel overwhelming for researchers in the early stages of development. That is where mISPL comes in. It is a streamlined, accessible "mini"…

Sep252026 by Heather WehnerNo Comments

New Pub: Implementation of AI-based Feedback and Assessment – IMPACT Project Evaluation Report

Artificial Intelligence, Feedback, Higher Education, Learning Analytics, Publication, Report

How can universities provide individual feedback and useful support systems to every student when class sizes continue to grow and learner needs become increasingly diverse? Artificial intelligence offers promising opportunities to provide personalized learning and teaching support on a larger scale. Yet as regulations regarding data protection, transparency and the responsible use of AI continue to evolve, institutions face a difficult balancing act between innovation and regulatory compliance. These issues were addressed in the IMPACT project (Implementation of AI-based Feedback and Assessment with Trusted Learning Analytics in Higher Education Institutions), which explored how AI can support students and educators throughout the academic journey. Its goal was to develop AI-based methods for the semi-automated analysis of academic texts, to allow for more effective support in learning, teaching and assessment. A key…

Aug172026 by Prof. Dr. Hendrik DrachslerNo Comments

Invited Talk at the NEES Center, Maceió, Brasil

Higher Education, Invited talk, Keynote, Keynote, Learning Analytics, Learning Design

In August, Hendrik Drachsler was invited to give a talk at the NEES research centre of the Federal University of Alagoas (UFAL) in Maceió, hosted by Prof. Dr Diego Dermeval. The lecture, "Design, Development and Evaluation of Highly Informative Learning Analytics", was organised together with the PPGI graduate programme and drew a full house of 80 participants. Hendrik's argument was straightforward. Most learning analytics is not wrong; it is just not informative enough. Counting clicks and minutes tells a teacher that something happened, not what to do about it. Moving from descriptive to truly informative analytics takes three things: indicators grounded in learning theory, a design process that includes the educators who will use the results, and a governance framework that makes the whole thing defensible. He presented the tools…

Jul312026 by Longwei CongNo Comments

[New Pub] Estimating LLM Grading Ability and Response Difficulty in Automatic Short Answer Grading via Item Response Theory

Artificial Intelligence, New Pub, School

On July 4, 2026, Longwei Cong virtually presented the paper “Estimating LLM Grading Ability and Response Difficulty in Automatic Short Answer Grading via Item Response Theory” at the ACL SIGEDU 21st Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2026). The work investigates automatic short answer grading with large language models from a psychometric perspective. Instead of relying only on aggregate evaluation metrics such as accuracy or macro-F1, the study applies Item Response Theory to jointly estimate the grading ability of individual LLMs and the difficulty of student responses. The results show that LLM graders differ not only in their overall performance, but also in their robustness when evaluating more difficult responses. In addition, highly difficult responses were found to induce systematic grading errors, including a tendency…

Jul312026 by Longwei CongNo Comments

[New Pub] Confidence Estimation in Automatic Short Answer Grading with LLMs

Artificial Intelligence, New Pub, School

On June 30, Longwei Cong presented the work “Confidence Estimation in Automatic Short Answer Grading with LLMs” at the 27th International Conference on Artificial Intelligence in Education (AIED 2026) in Seoul. The work investigates how confidence estimates can make automatic short-answer grading with large language models more reliable in practice. In addition to commonly used model-based confidence signals, it also considers uncertainty originating from the data itself. In particular, aleatoric uncertainty is estimated from the semantic heterogeneity of student responses and combined with several model-side confidence signals. The study evaluates these uncertainty estimates with a focus on practical use: whether they can support selective grading and help identify responses for which automated grading is less reliable and human review would be preferable. The paper is available via the following link:…

Jul242026 by Heather WehnerNo Comments

New Pub: The Dependency Dilemma of Machine Learning Aids

Artificial Intelligence, Journal, New Pub

AI and machine-learning decision aids can be a great support for employees and help them improve their work performance. An overreliance on algorithmic recommendations, on the other hand, may result in a long-term reduction of the employees’ ability to develop and maintain their own decision-making skills. This could be especially problematic when such AI and machine-learning decision aids are not available. Instead of overly relying on such AI-based recommendations, organizations should train their employees to continue developing their critical thinking and decision-making skills. Hendrik Drachsler and his research partners address this problematic dependency on AI tools and the conflict of interests in the workplace in their newly published article “The Dependency Dilemma: How Machine Learning Decision Aids can Undermine Skill Growth”. Using a controlled experiment, the authors found that participants…