Our Research Projects

“Don’t value what we measures -> measure what we value.”Bob Mislevey

DELTA: s   The realization of the DELTA project (Towards Digital Education with modern Learning Technologies and Assessment approaches) takes place in 5 stages:
  1. Writing personas: creation of so-called “personas” which reflect key users in the future of digital education at the Goethe University. Personas are developed on the basis of different empirical data – to this end, survey and structured interviews will be used.
  2. Expert group concept mapping study: An expert GCM study will be run to elicit success criteria for the development of a digital education infrastructure at the Goethe University.
  3. Innovation workshops: organisation of innovation workshops involving key actors at the Goethe University, external research institutes, the state ministry of education in Hesse and international experts, to formulate a plan for digital education at the Goethe University.
  4. Fruits & challenge workshops: organisation of workshops for the identification of so-called “low hanging fruits“, that is quickly accessible opportunities and long-term challenges to digital education at the Goethe University.
  5. DELTA report and conference: organisation of a wrap-up conference which summarises project outcomes and necessary activities for achieving DELTA plan objectives by 2025. The DELTA report will be handed to the President of the Goethe University at the conference.
Literature: Biedermann, D., Kalbfell, L., Schneider, J., Drachsler, H. (2019). Stakeholder attitudes towards digitalization in higher education institutions. DeLFI 2019 – Die 17. Fachtagung Bildungstechnologien der Fachgruppe Bildungstechnologien der gi e.V. Humboldt Universität zu Berlin, 16.-19. September 2019. Funding: Freunden und Förderer der Goethe-Universität Frankfurt
EduArc:

Mind map - Essential functions s
 
Digital Education Architecture Open learning resources in distributed learning Infrastructures - EduArc

State of the art In order to realize the potential of digitization for higher education, a cross-university digital ecosystem is needed that provides digital educational resources for distributed use (see M Kerres & Heinen, 2015). The BMBF feasibility study by Blees et al. (2016) on the infrastructure for open educational resources has shown for the higher education sector: There is an increasing amount of digital content on learning platforms and there are technologies available to provide this via repositories. What is needed is a solution that is based on a networked, federated infrastructure of decentralized repositories (see Heinen al., 2016) and that makes targeted use of open educational resources. Digital educational architectures must thus relate to the digital research infrastructure, in particular to the infrastructures for literary and research data and research data management, and the associated developments in the scientific information and librarianship (under the keyword "open science"). take into account (see Siegfried, 2017). For example, the Information Infrastructure Council (RfII) has issued recommendations to the GWK on the development of a national research data infrastructure. Since 2016, the DFG has been funding the Generic Research Data Infrastructure (GeRDI) project, which is developing a model for a distributed infrastructure. A federal infrastructure for research data in education is funded by the BMBF project "Verbund Forschungsdaten Bildung (VFDB)". At European level, the High Level Expert Group on the European Science Cloud (EOSC) deals with similar issues. Furthermore, the GOFAIR initiative aims to treat research data fairly, which was initiated by the BMBF and the Ministry of Science of the Netherlands. Similar to BW, Berlin and other federal states, the Digital University NRW has commissioned the planning of a subsidized infrastructure, through which open educational resources of the universities are made available and distributed.
 
Project overview
The project is developing a proven design concept for distributed learning infrastructures that will federate digital educational resources and other study-related information. It explores the technical, didactic and organizational conditions of an educational architecture that results from the networking of the digital infrastructure of universities and the interaction of state, public and private actors. It brings together distributed systems through open standards and interfaces, and is open to integrate future content providers and users. The announced architecture is to be interpreted for open as well as not openly licensed (references to) resources, because in teaching also perspectively different license variants will be relevant and expedient.
The project focuses on the challenges posed by the dissemination of openly licensed educational resources (OER / CC licenses) in an "informationally open ecosystem". Depending on the license, these can be used, commented on and edited by teachers and students free of charge , mixed and made available again on the net, which opens up special didactic opportunities for higher education (Heinen et al., 2016). At the same time there are technical-conceptual challenges in the provision of OER, but also in access to distributed repositories, especially in dealing with edits and versions, the return of user-generated data to the author of the OER, and the resulting quality mechanisms for reputation acquisition ( for authors) and quality assurance (for resources).
 
Methodical approach and interdisciplinary cooperation
The project pursues a design-oriented research approach in which design concepts are developed with prototypes and field trials. The type and intensity of use is recorded on the basis of (anonymised) objective data (behavioral traces), in addition to online questionnaires and guideline-based interviews with which the various groups of users are asked about their experiences and assessments.
First of all, the universities represented by the actors function as pilot universities. In addition to the spatial and organizational proximity to the applicants, predecessor projects of the applicants to the universities provide infrastructures in various stages of development, which can be used as the basis for a distributed learning infrastructure. On the basis of preparatory work, a dataset is defined that defines success parameters at various levels (technical, didactic, organizational) in order to be able to record the impact of the project and to verify the achievement of the objectives.
 
Literature
  1. Blees, Ingo, Hirschmann, Doris, Kühnlenz, Axel, Rittberger, Marc, Schulte, Jolika, Cohen, Nadia, … Khenkitisack, Phoutsada. (2016). Machbarkeitsstudie zum Aufbau und Betrieb von OER-Infrastrukturen in der Bildung. Frankfurt: Deutsches Institut für Internationale Pädagogische Forschung.
  2. Drachsler, H., Verbert, K., Santos, O. C., & Manouselis, N. (2015). Panorama of Recommender Systems to Support Learning. In F. Ricci, L. Rokach, & B. Shapira (Hrsg.), Recommender Systems Handbook (S. 421–451). Springer, Boston, MA.
  3. Euler, D., Hasanbegovic, J., Kerres, M., & Seufert, S. (2006). Handbuch der Kompetenzentwicklung für eLearning Innovationen: Eine Handlungsorientierung für innovative Bildungsarbeit in der Hochschule. Bern: Huber.
  4. Kerres, M. (2015). E-Learning vs. Digitalisierung der Bildung: Neues Label oder neues Paradigma? In A. Hohenstein & K. Wilbers (Hrsg.), Handbuch E-Learning. Köln: Deutscher Wirtschaftsdienst.
  5. Kerres, M., Getto, B., & Kunzendorf, M. (2010). RuhrCampusOnline: Strategische Hochschulkooperation in der Universitätsallianz Metropole Ruhr. Zeitschrift für Hochschulentwicklung, 5.
  6. Kerres, M., & Heinen, R. (2015). Open informational ecosystems: The missing link for sharing resources for education. Interna- tional Review of Research in Open and Distributed Learning, 16.
Funding: BMBF
DIFA: s
 
Education in a digitalized world opens the door for observing actual learning behavior with a fine-grained resolution. The proposed network of excellence investigates what is needed to release “computational psychometrics” (cf. von Davier, 2017) for formative assessment. We therefore will investigate how online-trace data as compared to standardized psychometric measures can be used to shed light on the learner’s knowledge, skills, and attributes that are in operation when using digital learning environments in higher education. For instance, timing behavior may be a useful indicator to evaluate individual learning engagement (cf. Nguyen, Huptych, & Rienties, 2018), and learning trajectories (e.g., semantic coherence of selected texts) may reflect self-regulation (see also Aleven, Roll, McLaren, & Koedinger, 2010; Winne, 2017).
 
For this purpose, the pedagogy concept and the design of the learning environment need to be closely aligned. That means that after specifying the learning outcomes of a course, we do not only need to define the assessment thereafter, but also need to think about potential learning analytics indicators that provide valuable insights into the state of the learner (Schneider et al., 2018). For this, the student’s learning process is inherently tied to the use of the learning environment for receiving, producing and exchanging information. This allows observing meaningful learning behavior that can be used to make inferences about the learner status and for providing feedback (formative assessment). To validate inferences based on learning analytics indicators well-proven classical standardized assessment instruments can be used as criteria. Thereby the DiFA project integrates methodological perspectives from educational assessment and learning analytics to develop new kinds of non-invasive assessment (“stealth assessment”, Shute, 2011) by exploiting digital trace data. Findings of this research are highly relevant for a better understanding learner’s behavior, learning outcomes and for providing individual feedback automatically to learners in digital environments.
 
The educational context of proposed project is initial and further teacher education. Within the DiFA project, we aim to develop an online qualification program for German teachers (students and professionals) on digital education. We can therefore reuse various learning materials that the consortium developed over the past years. This online qualification program will include various interactive online learning opportunities that will produce vast amounts of digital trace data.
 
Funding: Leibniz Kooperative Excellence
Die COVID-19 Pandemie hat gezeigt, dass sich viele Lehr-/Lerntätigkeiten ohne physische Präsenz durchführen lassen. Für psychomotorische Fähigkeiten gilt das kaum: ihre Entwicklung, wie sie in vielen Disziplinen notwendig sind (z.B. Medizin, Ingenieurwesen, Chemie, künstlerische Tätigkeiten, Sport) erfordert praktische Übung, direktes Feedback und Reflexion. Um gewünschte Lernerfolge zu erzielen, sind daher personeller Betreuungsaufwand und Materialeinsatz unabdingbar. Beides erhöht die Kosten und limitiert die Skalierungsmöglichkeiten der betroffenen Studiengänge: Experten sind rar und teuer, Materialeinsatz verursacht weitere Kosten. Aktuelle technologische Entwicklungen verändern diese Situation:
  • Mixed, augmented und virtual reality ermöglichen es, immersive Lern- und Übungsräume zu schaffen
  • Moderne Sensortechnologien können feingranular Bewegungen nachverfolgen und aufzeichnen
  • Big Data Methoden und ihre Anwendung in Learning Analytics können große Datenmengen analysieren und auswerten, was insbesondere bei datenintensiven Lernvorgängen, wie der Echtzeit-Analyse psychomotorischer Fähigkeiten unabdingbar ist
  • Maschinelle Lernverfahren (z.B. Reinforcement/Deep Learning) und generative Verfahren der künstlichen Intelligenz (z.B. generative adversarial networks) können große Datenmengen interpretieren, schlussfolgern und individuelles Feedback generieren
Bisher werden diese Technologien weitgehend getrennt betrachtet. MILKI-PSY hat zum Ziel, KI-gestützte, datenintensive, multimodale, immersive Lernumgebungen für das selbstständige Erlernen psychomotorischer Fähigkeiten zu schaffen. Dabei entsteht ein domänen-übergreifender Ansatz, der es ermöglicht, die Tätigkeiten von Experten multimodal aufzuzeichnen und diese Aufzeichnungen als Blaupausen für Lernende zu verwenden. Mit Hilfe KI-gestützter Analysen soll dabei der Lernfortschritt durch automatisierte Fehlererkennung und generiertes, individuelles Feedback unterstützt werden. So entstehen ganzheitliche, innovative Lernumgebungen für das Erlernen psychomotorischer Fähigkeiten, in denen personalisierte, KI-gestützte Lernunterstützung individuelle Lernprozesse auf Basis komplexer Datenanalysen ermöglicht.
Das HIKOF-DL Projekt hat zum Ziel Hochschulen und Firmen dabei zu unterstützen, das Potential der durch die COVID-19-Pandemie rasant gewachsenen Datenbestände der Online-Lehre mittels Künstlicher Intelligenz (KI) zu analysieren und in hoch informatives und kompetenzorientiertes Feedback für Studierende umzuwandeln. Dabei werden nicht grundständig neue Interventionen erforscht, sondern vielmehr etablierte Open-Source-Softwarelösungen von der internationalen Learning Analytics-Community sowie aus erfolgreichen BMBF-Projekten miteinander kombiniert und in Anwendung gebracht. Im Rahmen des Projektes wird ein projektbegleitender Beirat berufen, der die Anwendbarkeit der Technologien und des Wissens in der Praxis supervidier

Technologie und Innovation
Das Projekt baut auf der reichhaltigen Expertise und Unterstützung der Society of Learning Analytics Research (SoLAR) auf und verwendet erprobte Regelungen, Methoden und etablierte Open-Source-Softwarelösungen. Diese werden mit aktuellen Methoden und digitalen Werkzeugen der psychometrischen Kompetenzdiagnostik verknüpft. Konkret wird die international etablierte Learning Analytics-Software OnTask mit der in Deutschland entwickelten KAT-HS-Assessment- Software zu einem neuen wirkungsmächtigen Open Source Tool für hoch informatives und kompetenzorientiertes formatives wie summatives Feedback kombiniert.

Anwendungsbereich und Nutzen der neuen Technologie
Das HIKOF-DL Projekt legt seinen Fokus auf die Weiterentwicklung bestehender Interventionen zur Verbesserung der Online-Lehre und deren Transfer in die Lehr- und Lernpraxis. Perspektivisch wird eine langfristige Etablierung von KI in der Aus- und Weiterbildung angestrebt. Das neu entstehende Feedback-System wird an der Goethe-Universität durch die Partner GU-SD und DIPF implementiert und mit dem Partner GU-PSY in einer der größten und hinsichtlich der Teilnehmenden heterogensten Vorlesungen der Universität mit rund 1000 Studierenden evaluiert. Die Ergebnisse werden zweimal jährlich dem Beirat aus verschieden Unternehmen vorgestellt und auf dessen Anwendbarkeit im wirtschaftlichen Bereich bewertet.
AR4STE(A)M: Nowadays, competence requirements have changed with more jobs being subject to automation, technologies playing a bigger role in all areas of work and life, and entrepreneurial, social and civic competences becoming more relevant in order to ensure resilience and ability to adapt to change. Digitization affects how people live, interact, study and work. This makes investing in one’s digital skills throughout life of the utmost importance. As a consequence, education systems needs to adapt to this reality. Innovation is ‘extremely relevant’ for meeting the education sector’s needs and digital technology has huge, untapped potential for improving education, especially in STEM subjects.

With the AR4STE(A)M project the Educational Technologies Team tackles these Issue by leading a strategic partnership supported by the European Commission in order to raise awareness on the importance of choosing STE(A)M studies for pursuing successful careers, especially among young students. Over the course of 2, 5 years the Team will work – together with 6 Partners from 6 EU countries – to give school teachers innovative instruments and tools to overcome the traditional method of teaching and learning engaging students to learn while enjoying. Supporting educational institutions in integrating immersive technologies and GBL for teaching STE(A)M within their curricula.

  Funding: Erasmus+

Basierend auf ersten Forschungsergebnissen rund um das SERNE -Selfregulation Widget haben wir zusammen mit Prof. Dr. Garvin Brod von der DIPF Abteilung Bildung und Entwicklung ein Projekt zur Unterstützung von selbstregulierten Lernenprozessen für Schüler*Innen im HomeSchooling einwerben können. Das PROMPT Projekt wird durch das Distr@l – Förderprogramm des Hessische Ministerium für Digitale Strategie und Entwicklung gefördert.

by Wanqij, CC BY-SA 4.0 via Wikimedia Commons
by Wanqij, CC BY-SA 4.0 via Wikimedia Commons

Unterstützung des Selbstreguliertes Lernen fällt insbesondere jüngeren Schülerinnen und Schülern schwer. Das Projekt PROMPT hat das Ziel, die Erkenntnisse aus der Forschung zur Verbesserung von selbstreguliertem Lernen in digitalen Lernumgebungen in die Anwendung bei Schulkindern zu überführen. Dieses Wissen soll möglichst breit zugänglich gemacht werden, sodass es Unternehmen der Bildungswirtschaft und Bildungsinstitutionen unmittelbar für neue Produkte oder zur Verbesserung bestehender Anwendungen nutzen können. Einen zentralen Aspekt dieses Wissenstransfers soll ein Prototyp einer kindgerechten Lernplanungs-App darstellen. Dieser Prototyp soll die wissenschaftlichen Erkenntnisse, die für eine Verbesserung des selbstregulierten Lernens mit Lern-Apps bei Schulkindern wichtig sind, praktisch umsetzen und illustrieren. Anbieter und Nutzer jeglicher Lern-Apps können auf dieses Instrument zurückgreifen und für ihre Zwecke adaptieren. Im Projekt erfolgt eine ausführliche, mehrschrittige Optimierungsforschung zur Gestaltung der Lernplanungs-App, sodass diese größtmögliche Effektivität besitzt. Diese Optimierung besteht zunächst aus einer Reihe von kleineren Randomized Controlled Trials (RCTs), in denen die verschiedenen inhaltlichen Komponenten der App (Lernzielformulierung, Lernzielmotivation, Wenn-Dann-Plan zum Umgang mit Hindernissen bei der Zielerreichung, Überwachung des Lernfortschritts) systematisch variiert werden, um zu überprüfen, welche Kombination von Komponenten die größtmögliche Effektivität hat. Die vielversprechendsten Kombinationen werden im Anschluss in einem großen RCT in einer repräsentativen Stichprobe überprüft, um die Effektivität des Prototyps wissenschaftlich belastbar zu validieren. Abschließend wird die finale Version des Prototyps für die breite Nutzbarkeit und Zugänglichkeit für Unternehmen und Bildungsinstitutionen aufbereitet und als Open Educational Resource verbreitet.

Basierend auf ersten Forschungsergebnissen rund um die Trusted Learning Analytics Infrastructure der EduTec Gruppe, haben wir zusammen mit Prof. Dr. Knut Neumann from Leibniz-Institut für die Pädagogik der Naturwissenschaften und Mathematik (IPN) in Kiel und Prof. Dr. Nikol Rummel von der Ruhr-Universität Bochum ein Projekt zur Analyse und Förderung von Lernverläufen zur Entwicklung von Kompetenzen (AFLEK) einwerben können. Das AFLEK Projekt wird durch das BMBF Programm Digitalisierung II - Forschung zur Gestaltung von Bildungsprozessen unter den Bedingungen des digitalen Wandels gefördert.

Digitalen Technologien wird ein hohes Potential für die Optimierung von Bildungsprozessen zugesprochen. Sie erlauben ein stärker personalisiertes Lernen, das es allen Schülerinnen und Schülern ermöglicht, die für eine gesellschaftliche und insbesondere berufliche Teilhabe erforderlichen Kompetenzen zu erwerben. Dies setzt voraus, dass Lernverläufe die nicht zu Kompetenzentwicklung (unproductive learning; vgl. Kapur, 2016) führen zeitnah erkannt werden und entsprechend darauf reagiert wird. Dies für alle Schülerinnen und Schüler zu leisten, ist für Lehrkräfte im herkömmlichen Unterricht nur schwer bis gar nicht möglich. Im Projekt „Learning Progression Analytics“ (LPA) sollen Erkenntnisse darüber gewonnen werden, wie die umfangreichen bei der Bearbeitung digitaler Lerneinheiten generierten Daten zu fachinhaltlichen Lernprozessen automatisch ausgewertet und der Lehrkraft in nahezu Echtzeit zur Verfügung gestellt werden können, um unproduktive Lernverläufe zeitnah zu erkennen, die ursächlichen Lernschwierigkeiten zu identifizieren und instruktionale Interventionen zu unterstützen, die die unproduktiven Lernverläufe in produktive überführen. Damit soll die Grundlage für die Entwicklung von Assistenzsystemen für Schülerinnen und Schüler bzw. Lehrkräfte und damit für personalisiertes Lernen im Schulunterricht geschaffen werden.

The Coronavirus Crisis has spurred an urgent need to support students’ learning via digital technologies. This need highlights the importance of efforts in the educational sector to take advantage of the unique possibilities to promote learning based upon digital technologies and the analysis of digital data. Well beyond using digital platforms for distributing tasks to students, digital technologies make it possible to track individual students’ learning and provide targeted support tailored to the individual needs of each student. A stronger individualization of learning has been advocated as a means to support all students in developing the competence required for occupational, societal, and cultural participation – in particular in such critical domains as mathematics and science (e.g. Reinhold et al., 2018; Schiepe-Tiska, Rönnebeck & Neumann, 2018). However, individualized learning, also known as personalized and adaptive learning (Brusilovsky, 2012), requires the continuous evaluation of students’ learning, reconstruction of students’ learning trajectories and extrapolation of these trajectories with respect to students’ competence development (Hattie & Timperley, 2007; Kapur, 2016; VanLehn, 2006). This requires:
  1. a theory of learning, and based thereupon,
  2. a model of competence development and
  3. methodologies that allow for a continuous assessment of students’ learning over sequences of learning activities and subsequent mapping to students’ competence development.
Since digital technologies have become increasingly ubiquitous in mathematics and science classrooms, they may lend themselves to developing such methodology. As students are working with digital technologies, their interactions with the technologies can be recorded and automatically analyzed. Automated analysis of these interactions allows for a just-in-time assessment of individual students’ performance (e.g. Gobert et al., 2013). Drawing on these advances, we aim to investigate the extent to which data originating from students interactions with digital technologies in mathematics and science classrooms can be used to 1) continuously evaluate individual students’ learning, 2) reconstruct trajectories in their learning over sequences of learning activities and 3) identify those trajectories that align with the development of competence in mathematics and science. In doing so, we aim to provide the theoretical and methodological foundations for a stronger individualization of mathematics and science education to support all students in developing the competence needed for societal, cultural and occupational participation and thus meeting educational goals.
The DELTA School is intended to develop an Ecosystem for Digital Education as well as facilitate scientific research within this EcoSystem. The DELTA school is supposed to apply an Inquiry-Based Learning model, where the students apply EduTec interventions themselves to conduct research, as well receive latest insights from this research on digital education in their and are made tangible for the students. The teaching, learning and assessment methods in the DELTA School not only aim to build up a body of knowledge, but also to promote competence development and enable learners to work creatively and problem-solving with acquired knowledge. Methods of Inquiry Based Learning lend themselves to this. Inquiry-based learning, which goes back to constructivist learning theories, tries to create learning settings in which learners can gather personal and authentic experiences and derive meanings and explanations from them. Especially collaborative competence, which plays a central role in the interdisciplinary field of Educatioanl Technologies, is promoted by the methods described below. Students are confronted with an EduTec problem in the field of education. In student groups, they derive topic-specific learning questions, define knowledge gaps and work out solutions independently. The student groups are accompanied by the lecturer, who supports the formulation of suitable learning questions and can intervene in the derived solutions if necessary. The aim is to further develop the students' critical reflection skills as well as problem-solving and social competence and to open up an understanding of interdisciplinary approaches. The DELTA School is a overarching initiative that aims to combine digital education at Goethe University, in combination with those at the DIPF and studiumdigitale. The DELTA School can draw on a wide range of preliminary work, as well as many prototypes and expertise on the part of the DIPF (EduTec, EduCS, TBA) and a competence model for the development of digital key skills from studiumDigitale, as well as existing events from the Master's specialisation Educational Technologies of the Department 12 Mathematics / Computer Science.
Die Erforschung von Faktoren, die einen Studienerfolg und -abbruch in digitalen Studienformaten beeinflussen, wurde sowohl im nationalen als auch im internationalen Umfeld bislang vernachlässigt. Der DZHW-Bericht „Zwischen Studienerwartungen und Studienwirklichkeit“ (Heublein et al., 2017) dokumentiert Studienerfolg und -abbruch beeinflussende Faktoren sowie deren Wechselwirkungen insbesondere für traditionelle Studienformate an Präsenzhochschulen. Die Datengrundlage berücksichtigt keine Merkmale, die sich auf digitale Studienformate in der Fernlehre beziehen. Fernhochschulen bilden aufgrund ihres durchgängig hohen Anteils an digitalen Lehr-/Lernformaten ein ideales Feld, um eine Vergleichsgruppe zum Präsenzstudium zu erschließen und neues Wissen über die Wirksamkeit digitaler Hochschulbildung in Hinblick auf Studienerfolg zu generieren. Abbruchentscheidungen sind stets multikausal zu betrachten (vgl. Heublein 2010 in Hillebrecht 2019, S. 67; Neugebauer et al., 2019; Röwert et al., 2017), so dass sowohl die subjektive als auch die curriculare und institutionelle Ebene als Untersuchungsgrundlage betrachtet werden muss. Gemäß Themencluster A werden Faktoren, die den Studienerfolg in digitalen Studienformaten beeinflussen, vergleichbar zu und auf der Grundlage der DZHW-Studie (Heublein et al., 2017) sowie ihres Untersuchungsdesigns und dessen Erweiterung analysiert und damit die forschungsleitende Fragestellung, welche Faktoren Studienerfolg und -abbruch in der digitalen Fernlehre zum Präsenzstudium in einem bundesweiten Vergleich beeinflussen, beantwortet. Konkret geschieht dies anhand folgender Fragen: 1. Wie gestalten sich die Einflussfaktoren auf den subjektiven, curricularen und institutionellen Ebenen für den Studienerfolg bzw. -abbruch in digitalen Studienformaten? 2. Welche Ausprägungen und Wechselwirkungen lassen sich empirisch ermitteln? 3. Wie lässt sich ein Vergleich der zwei Studienformate (digital und in Präsenz) anhand der Faktorenanalyse ziehen? 4. Wie stellen sich die Wechselwirkungen zwischen diesen Faktoren in digitalen Studienformaten dar?
The European strategic project in higher education titled RAPIDE- Relevant assessment and pedagogies for inclusive digital education starts on Monday, March 1st 2021. The project, coordinated by the Faculty of Organization and Informatics (FOI) of the University of Zagreb, is the 24-month project funded under the Erasmus + program of the European Commission in the total amount of 219,085 euros. The goal of this  project is to co-create, implement and share innovative pedagogies and aligned assessment for relevant and inclusive digital education in order to deal with the COVID-19 induced and similar crises and to support meaningful digital transformation of higher education in (post)COVID era. The kick-off meeting of the project team including the coordinator and project partners - School of Medicine of the University of Zagreb, University of Rijeka, Open University, Delft University of Technology and Goethe University, is scheduled for April 2021. During the project, the partners will work on 4 main results, and will start by creating an Open educational resources and e-course for flipped classroom (FC) and work based learning (WBL) for the use in an online environment with the main aim to provide teachers and students with an original resources designed in a form of research-based practical guidelines for FC and WBL approaches in an online environment and as an open e-course. Further, the project partners will work on the development of the Toolkit for assessment of students in FC and WBL with the main aim to provide HE teachers, Dashboard model that supports inclusive FC and WBL and Code of practice on impact analysis of innovative pedagogies. One of the biggest expected impacts of the project will be to advance the included institutions through joint work enabling its teaching staff and students to experience the newest innovative (post)COVID world trends in hybrid teaching environments. More information on the expected results will be available soon on the official project website: https://rapide-project.eu
With almost 10 million euros in funding, the Volkswagen Foundation encourages research projects that explore how artificial intelligence will affect society. Goethe University Frankfurt (GU) was successful with an application that looks at developments in human-machine interaction in education. “From Machine Learning to Machine Teaching (ML2MT) – Making Machines AND Humans Smarter” – this is the title of the project that the economist Prof. Oliver Hinz applied for in an interdisciplinary project together with colleagues from various subjects. The success of learning machines, as in the prime example of the board game Go (in the computer version “AlphaGo Zero”), has inspired scientists. Their project aims at a better understanding of how humans and machines in collaborative human-AI systems can develop new knowledge in symbiotic interaction with each other. To this end, the consortium is researching the analytical and technical foundations that are responsible for the successful transfer of new knowledge from intelligent machines to humans and vice versa. This is being investigated by means of hybrid human-machine systems in case of studies from medical diagnostics, economic decision-making and financial market forecasting. The team wants to derive generalisable socio-technological and psychological findings and make recommendations to further improve the interaction between humans and machines. The individual members of the project are: Prof. Oliver Hinz (Economics, GU (lead)), Prof. Yee Lee Shing (Developmental Psychology, GU), Prof. Loriana Pelizzon (Economics, GU) and Prof. Tobias Tröger (Law, GU, both also at the Leibniz Institute for Financial Market Research SAFE, Frankfurt), Prof. Gernot Rohde (University Hospital Frankfurt/Main and GU), Prof. Kristian Kersting (Computer Science, TU Darmstadt), & Prof. Hendrik Drachsler (Computer Science, GU, and DIPF | Leibniz Institute for Human Development and Educational Information, Frankfurt/Main location). The Volkswagen Foundation is funding seven project consortia from the social and technical sciences with a total of 9.8 million euros. With its “Artificial Intelligence” initiative, it aims to promote interdisciplinary and transnational research on the responsible further development of AI systems. “The newly approved projects focus on areas in which AI systems are already being used or will be used in the near future, for example in medical diagnostics or preventive remote therapy, but also in financial market forecasting, scientific image processing or journalism,” says Dr Henrike Hartmann, head of the funding department. “The researchers are thinking one step ahead, anticipating the impact of AI on society and how to make it positive.” All selected projects are scheduled to run for four years and will each receive around 1.5 million euros in funding. The initiative “Artificial Intelligence – Its Impact on Tomorrow’s Society” has been running since 2017, and a total of 33.9 million euros has been approved to date. The content of the initiative will be further developed in 2022. Further information on the Volkswagen Foundation’s initiative “Artificial Intelligence – Its Impact on Tomorrow’s Society” can be found at www.volkswagenstiftung.de/kuenstliche-intelligenz.
The joint project IMPACT promotes the improvement of higher education through the scalable use of artificial intelligence (AI) methods for the (partially) automated analysis of texts. Students are to receive so-called "highly informative and personalized feedback" (HIF) at the various stages of their studies - as prospective students and new students, during the course of their studies and at the completion of academic achievements. This helps to cope with paracurricular study requirements and to promote individual learning goals as well as self-regulation strategies for future learning. In addition, uncertainty and excessive demands on students can be reduced, teachers can be relieved, and active engagement with AI-generated feedback by students and teachers can be promoted. In the IMPACT project, open-source software solutions and preliminary work already established at the participating universities (Goethe University, Humboldt University Berlin, Hagen University, Freie Universität Berlin, University of Bremen) will be combined and put to use. In the process, rapidly growing data sets can be analyzed using AI and converted into personalized HIF. For example, students can be given low-threshold advice as needed in the Student Orientation and Entrance Period (SOEP); they will benefit to a far greater extent from the volumes of data analyzed in advance than would be possible in the form of personal support. It is expected that chatbots can be helpful support systems in the future and can be relevant information about the respective university, as well as about the respective student topic. AI can be used to provide feedback, especially in large courses, faster and to more students than can be provided by a lecturer and corresponding tutors. AI can also be very helpful in evaluating exams and saving time. In the IMPACT project, we will explore to what extent, for example, free-text entries in exams can be classified by AI. Free-text assignments are a popular way for many students to express what they've learned, but grading them is very time-consuming and resource-intensive. AI applications could take some of the burden off instructors; learning processes would be supported beyond courses. Individual student support through Trusted Learning Analytics can lead to students becoming more engaged with learning objectives, requirements and assessment criteria, and developing an understanding themselves of what action strategies conducive to learning look like. https://blog.studiumdigitale.uni-frankfurt.de/impact/
Das Projekt ALI (AI and digital Technology in Learning and Instruction) aims to develop an interdisciplinary study program on the use of artificial intelligence and digital technologies in educational processes. At the Goethe University Frankfurt (GU) an accordingly denominated tenure-track professorship in educational psychology will be established. This will enable a permanent cooperation between the departments of computer science, psychology and teacher education as well as relevant subsidiary subjects at the GU in cooperation with existing work units of computer science on AI, university didactics, educational psychology, as well as the didactics of computer science. Based on numerous preliminary studies, the program is designed in such a way that students from the above-mentioned departments can apply for this new Master's program, but can also acquire individual modules from the start of the program for certificates or specializations appropriate to their degree. The program uses inquiry-based learning and integrates elements of AI-supported teaching and learning into its study practice. Based on theoretical approaches of empirical educational research, a considerable increase in effectiveness compared to conventional teaching can be expected (e.g. through AI-generated feedback to relieve the teachers). At the end of the program, an accredited, interdisciplinary master's degree program "AI and Digital Technologies in Learning and Instruction" as well as modular certificates for the selectable focal points in master's degree programs will be available. Additionally, continuing education and advanced training programs for academic professionals, especially in the education sector, will be offered.
Three-years collaborative research project between DIPF EduTec and Cologne Game Lab, starting from 1st September 2022. Artificial Intelligence (AI) systems can provide automatic, personalised and real-time feedback to learners in distance learning settings when a human expert is unavailable. AI feedback has the potential to be always available and can allow learners to practice deliberately and repeatedly at their own pace. Embedding AI feedback into immersive and multimodal technologies, like Augmented and Virtual Reality (AR/VR) or sensor-based systems, enables learners to train physical learning tasks alongside the more traditional cognitive tasks. We refer to these AI systems as Multimodal Tutors. One scenario that needs to be considered when replacing human feedback is that the AI system may deviate from the learner’s expectations and the teacher’s learning goals, risking becoming unethical. This challenge is known in AI ethics research as the “alignment problem” and refers to the importance of designing AI systems that align with human values. This project tackles the “alignment problem” in distance teaching in the particular context of presentation skills training. To tackle the problem, we design a Multimodal Tutor to train presentation skills such as body position, pauses, and voice intonation. The Multimodal Tutor features various game-oriented immersive scenarios on a mixed-reality headset to allow learners to practice autonomously, supported by AI-generated feedback. The human experts can monitor the learning process asynchronously via an Alignment Dashboard. They can assess the learner performance and the quality of the AI feedback, ensuring it is correctly aligned with the learning goals.

If you are interested in the products we have developed in the research projects navigate to the product page.