20. Research in Innovative Intercultural Learning Environments
Paper
Sh-AI-ping Future Research. A New AI-based Research-Approach in Examining Trajectories and Student Drop-out in Higher Education.
Désirée Wägerle, Augustin Kelava
University of Tübingen, Germany (Methods Center)
Presenting Author: Wägerle, Désirée
In a rapidly changing world, it is not only the higher education system that is facing an era of uncertainty, characterised by recent geopolitical issues, social divisions, a significant increase in knowledge and a growing scepticism towards science (Gassmann et al., 2023, p. 1). In this context, recently developed new digital technologies such as generative artificial intelligence [GenAI] has unquestionably marked a noteworthy historical occurrence, offering the potential to transform the entire higher education sector (Bannister et al., 2023, p. 402; Gassmann et al., 2023). GenAI can be seen as a technology that enables industrial systems and devices to learn independently, process information and make decisions without human intervention (Quy et al., 2023, pp. 4–5). In contrast to other AI techniques, GenAI can also generate new and original content such as text, images, videos, audio, and 3D models (Escotet, 2023, p. 3). The potential benefits and hopes that these new applications could offer for enhancing education and research are considerable, but the growth also leads to new ethical considerations and potential risks that especially higher education institutions will need to address in the future (Meyer et al., 2023, p. 1; Zawacki-Richter et al., 2019, pp. 1–2). Despite the rapid progress and increasing attention given to these new technologies in higher education, recent reviews have raised concerns about the low number of authors from education departments involved in research on them, underscoring the need for pedagogical research and research perspectives on these technological advances (Zawacki-Richter et al., 2019, p. 22). Existing research on GenAI in higher education often focuses on academic integrity, institutional regulations, plagiarism concerns in specific applications, the benefits of using AI for educational innovation or students' attitudes toward AI. However, only a few studies have explicitly examined the use of GenAI as a methodology in research. As a result, this aspect is highlighted as an area for future research (Bannister et al., 2023).
Building on this need, we present a recently in the context of higher educational research created, innovative approach to improve data collection and data analysis methods, integrating conventional and generative AI-assisted web-crawling techniques based on natural language models. Large Language Models [LLM] are a specific Natural Language Processing technology, trained on large text datasets derived from public and licensed third-party sources and are capable of producing human-like speech and performing a range of language processing tasks (Alqahtani et al., 2023). The project, in which we developed the method, aims to investigate patterns of social inequalities during the so-called postdoc phase, which is understood as the transitional phase from completing a doctorate to obtaining a first professorship (Krawietz et al., 2013; Roman, 2022). The analysis is based on specific data points extracted from the online profiles of all professors working in departments of Educational Science or Human Medicine at German state universities. These data include information about gender, age, number and type of career transitions, post-doctoral qualifications, and publication activity during the postdoc phase among a total sample size of N=7.204 professors. Until now, manual methods have been used to tackle the problem of different websites not having a standardised data display structure (Baader et al., 2017; Lutter et al., 2022). However, these manual methods are resource-intensive, time-consuming and error-prone. Our new AI-powered web crawling approach allows for the automated extraction and organization of crucial information from websites for research purposes. By this, our approach offers several advantages, primarily through its ability to automate processes and make them more efficient. Automation enables the collection of specific information from a variety of online sources, reducing time and effort while improving accuracy and efficiency.
Methodology, Methods, Research Instruments or Sources UsedIn our dataset, we had to extract the unstructured CV-data and relatively structured publication data from 7,204 people working at more than 100 universities, each of them comprising a large number of personal pages with different web structures. Since manual approaches of data collection require extensive resources, are very time-consuming and tend to cause inaccura-cies and errors (Arasu & Garcia-Molina, 2003, p. 338), our approach on data extraction and structuring is primarily based on the use of Web Data Scraping [WDS]. WDS refers to the pro-cess of extracting data from websites by using automated techniques. In this method, a com-puter program accesses a website's markup languages such as HTML to code and retrieves specific information from it. Traditional WDS techniques can handle various data formats and accurately extract specific data points. They are a fast and efficient method for extracting structured data (Parvez et al., 2018), which is why we used them to extract the publication data from the databases GoogleScholar and PubMed. In contrast to prestructured platforms, which provide profile data in a uniformly predefined, standardized structure, accessing, structuring and analysing information on university websites is significantly more complex (Arasu & Garcia-Molina, 2003). This can be attributed to various aspects. The websites within and be-tween universities present for example diverse formats, unstructured content with semantic heterogeneity, dynamic information that can change based on user interaction, and the webpages content sometimes contain noise or errors requiring cleansing. Since conventional WDS breaks in these cases (Parvez et al., 2018), we used a generative AI-assisted WDS-approach to extract the CV-data from the university-websites. We therefore used the NLM GPT-4, which is based on its predecessor GPT-3 (Brown et al., 2020, p. 5). Introduced in 2023 by OpenAI (Angelis et al., 2023, p. 1), GPT-4 is the latest version and currently the most powerful LLM (Hao et al., 2023, p. 10), providing the ability to process both textual and visual inputs and generate text-based outputs (Alqahtani et al., 2023, p. 1237; OpenAI et al., 2023, p. 1). One of the main issues we are facing to solve in using GPT-4 was and still is to handle its tendency to hallucinate, which means to create content that is nonsensical or untrue (Alqahtani et al., 2023, p. 1237; OpenAI et al., 2023, p. 68). The data generated in this manner was consolidated in a shared database and uniformly structured with AI-supported techniques.
Conclusions, Expected Outcomes or FindingsThe field of higher education is currently undergoing a process of transformation due to the emergence of GenAI techniques. Given the opportunities and risks associated with the use of such technologies, particularly in their application in higher education research, it is crucial for educational researchers to catch up and bring their experiences and perspectives to bear in exploring their application as an empirical methodology in academic discourse. Our innovative method of data collection and analysis was pursued with a specific educational policy goal in mind - namely, to investigate social inequalities within the German higher education system. Thus, our work can be seen as a step towards integrating an educational science perspective into the discourse on generative AI research methods. The approach of GenAI, specifically the NLM GPT-4 in combination with conventional WSD techniques developed within the project enables a more efficient and precise automated extraction and organisation of unstructured internet data. As our approach is one of the first, if not the very first of its kind, future research efforts should focus on further improving these techniques to enable greater accuracy and efficiency in the automatic extraction of structured web-based information. For example, what has been done manually in our approach so far is the search for relevant websites, for which a solution still needs to be found. Additionally challenging remains the handling of hallucinations of GPT-4. As briefly outlined, hallucination, also known more precisely as confabulation, de-scribes the generation of plausible but factually incorrect information by an AI model without intent to deceive (Alqahtani et al., 2023, p. 1237). Despite existing challenges, our innovative approach offers promising areas of application for future research in the field of higher educa-tion. The possible application to other empirical research scenarios could thus be a key focus of future considerations.
ReferencesAlqahtani, T. et al. (2023). The emergent role of artificial intelligence, natural learning pro-cessing, and large language models in higher education and research. Research in Social & Administrative Pharmacy: RSAP, 19(8), 1236–1242.
Arasu, A., Garcia-Molina, H. (2003). Extracting structured data from Web pages, 337–348.
Baader, M. et al. (2017). Equal opportunities in the post-doctoral phase in Germany? Europe-an Educational Research Journal, 16(2-3), 277–297.
Bannister, P. et al. (2023). A Systematic Review of Generative AI and (English Medium Instruc-tion) Higher Education. Aula Abierta, 52(4), 401–409.
Brown, T. et al. (2020). Language Models are Few-Shot Learners, 1–75. https://arxiv.org/abs/2005.14165
DeAngelis, L. et al. (2023). Chatgpt and the rise of large language models: The new AI-driven infodemic threat in public health. Frontiers in Public Health, 11, 1–8.
Escotet, M. Á. (2023). The optimistic future of Artificial Intelligence in higher education. PRO-SPECTS, 1–10.
Gassmann, O. et al. (2023). Universities in an age of uncertainty: 44 propositions on the future of universities.: [White Paper]. University of St. Gallen. Wissenschaftsmanagement, 21, 1–7. https://www.alexandria.unisg.ch/handle/20.500.14171/117981
Hao, Y. et al. (2023). E&V: Prompting Large Language Models to Perform Static Analysis by Pseudo-code Execution and Verification, 1–13.
Krawietz, J. et al. (2013). Übergänge in der Hochschule. In W. Schröer, et al. (Eds.), Hand-buch Übergänge (651-687). Beltz Juventa.
Lutter, M. et al. (2022). Gender differences in the determinants of becoming a professor in Germany. An event history analysis of academic psychologists from 1980 to 2019. Research Policy, 51(6).
Meyer, J. et al. (2023). Chatgpt and large language models in academia: Opportunities and challenges. BigData Mining, 16(1), 20.
OpenAI (2023). GPT-4 Technical Report, 1–100.
Parvez, M. et al. (2018). Analysis Of Different Web Data Extraction Techniques, 1–7.
Quy, V. et al. (2023). AI and Digital Transformation in Higher Education: Vision and Approach of a Specific University in Vietnam. Sustain-ability, 15(14), 1–16.
Roman, N. (2022). Honeymoon is over? Strategien im Umgang mit Selbstpositionierungen in der Postdocphase. In S. Korff & I. Truschkat (Eds.), Übergänge in Wissenschaftskarrieren (pp. 73–94). Springer Fachmedien Wiesbaden.
Zawacki-Richter, O., Marín, V., Bond, M., Gouverneur, F. (2019). Systematic review of re-search on artificial intelligence applications in higher education – where are the educators? International Journal of Educational Technology in Higher Education, 16(1), 1–27.
20. Research in Innovative Intercultural Learning Environments
Paper
Leading Innovation in Colleges of Education: Integration of Heutagogy Approach
Michal Ganz-Meishar1, Regina Benchetrit2
1Levinsky-Wingate Academic, Israel; 2Kaye College, Israel
Presenting Author: Ganz-Meishar, Michal;
Benchetrit, Regina
The heutagogy approach aims to overcome the crisis in Western world education and to change the contemporary education system, which faces Western society's multicultural, heterogeneous, dynamic, and evolving diversity. This is true for higher education institutions, where the prevailing teaching-learning paradigm no longer meets the learners' needs and society's demands. At the center of these requirements are self-management abilities, reflective and critical thinking, digital literacy, innovation, problem-solving, and collaboration and communication abilities (Blaschke, 2021).
Heutagogy, or self-directed learning, is an approach to learning and inquiry in which learners conduct self-inquiry. They determine what to study and explore in each content area. They decide how to do this, what sources of information they will base themselves on, what the results of their research will be, and how these will be presented to others. The learners wander between human knowledge spaces and Internet knowledge spaces when the purpose of this wandering movement is to satisfy their curiosity and bring them to understand the object of their research. Also, the learners are central partners in evaluating the research they have carried out themselves. They are the ones who determine whether and to what extent they have achieved the learning objectives (Glassner & Back, 2020; Hase & Kenyon, 2000).
The Self-inquiry journey of the heutagogy learners is integrally accompanied by self-thought, reflective and critical writing about the progress of their research, their attitude to their study, and teamwork within which the questions that interest them are investigated (Blaschke & Hase, 2021). The Heutagogy changes the known and accepted education orders.
In this study and investigation, the teachers or lecturers are no longer the primary sources of knowledge. Their traditional role changes, and they become mentors and advisors. Learning through heutagogy is no longer subject to a linear and uniform curriculum "imposed from above," to know and predetermined patterns, or exclusive reliance on academic information sources. The ways of learning are diverse, and there is no one way of knowledge suitable for all learners. Learning and reflection about learning bring learners to an understanding of their preferred learning style. This understanding will help them to continue learning throughout their lives, satisfy their curiosity, and strengthen their autonomy to choose what to learn and how (Moore, 2020; Blaschke, 2021).
Teacher training colleges are an effective anchor in the development of the teacher's professional personality and the formation of his image as an educator who knows how to integrate students from different cultures in a multicultural environment not out of paternalism of a majority group, but out of social solidarity, eradicating the feeling of foreignness and hostility and implementing teaching methods that mobilize personal capital (Butler & Milley, 2020; Ratnam, 2020).
The study presents insights regarding possible achievements, challenges, and changes required when implementing the heutagogy approach in two colleges of education in the center and one in south Israel.
The diversity in the number, age, education, and academic abilities of the students in each course, as well as the unique characteristics of the colleges and the differences in their disciplinary affiliations, allow for observing the heutagogy approach from a broad perspective.
Research questions:
1. How do the students and teacher-teachers who participated in these courses perceive the heutagogy approach?
2. What are the challenges your teachers faced in these courses?
3. What are the characteristics of the change required in teacher training so that it will be possible to incorporate learning in the way of heutagogy?
Methodology, Methods, Research Instruments or Sources UsedThe research approach is a case study in a qualitative constructivist centered on a categorical content analysis of the four heutagogy courses taught in four teaching colleges in Israel. The analysis examined the achievements, challenges, and actions of the investigated cases (Merriam, 2009; Adler & Adler, 1994; Kawulich, 2005).
The research tools include the students' learning diaries, the lecturers' written responses to the learning diaries, correspondence between the learners and the lecturers, and reflections written by the learners at the end of the course. The research's use of these sources received the approval of an institutional ethics committee, and all the students whose diaries and correspondence were studied approved their use for the research.
In the process of analyzing and determining the categories, the course researchers - based themselves on an inductive approach in which the texts were divided into units of analysis and through a "constant comparison" (Corbin & Strauss, 2008) of the various branches of research, we looked for similarities, differences, and connections between the diverse references. Thus, in the process of conceptualization (coding), we identified themes and concepts that were gathered into categories and subcategories, through which we sought to find meanings that lie in the course data and relate to the answers to the research questions (Creswell et al., 2018).
The participants were asked to create interest groups according to the chosen research topic derived from the course topic. They researched topics they chose for learning in pairs or groups and set common goals to maintain the research quality. They decide how to present the study and evaluate themselves according to the criteria while building an evaluation scale. Each course lecturer had given regular and continuous guidance and accompaniment. The four courses: (1) Leadership, policy, and Organization in Early Childhood Education Systems took place in the second and last year of a hybrid program for early childhood education as part of the 33 students' master's degree curriculum at a religious state college. (2) English online course: "Learn to write, write to learn." 13 students in their second year of studies majoring in English to improve their written expression skills; (3) "New teacher" courses in primary and secondary schools - 62 teachers participated in face-to-face learning and at Zoom; (4) Education course for social activism - an asynchronous course for 64 students within the framework of academic retraining for teaching.
Conclusions, Expected Outcomes or FindingsThis research offers a unique and innovative approach to assimilating the heutagogy approach as a teaching method adapted to the learner and lifelong learning. The lecturers and learners must be active references (Glassner & Back, 2020). According to the heutagogy approach, the lecturers should provide detailed explanations, personal conversations, close monitoring of the learning progress, and a quick response to write in the "journey diaries."These may help the facilitators calm the learners and bring them to cooperation and productivity already in the first stages of the courses.
There is a change in the role of the lecturers in these courses to assimilate the heutagogy. They are asked to have an open and in-depth institutional dialogue that will allow them to consult and share. This approach encourages the lecturers to think personally about their personal and professional identity and the inevitable change in their role and status in the educational institution.
The lecturers have shown satisfaction with teaching according to the heutagogy approach, and they are interested in continuing this way and even expanding it to additional courses in their colleges. By implementing this way of learning, they seek to propose an innovative change in which the learner also has an appropriate place. However, it should be noted that the realization method differs in the degree of freedom and independence students have been allowed in each course. One student wrote: "I hope that this approach will be able to enter other places in the academy, into other courses where it can be integrated and will awaken in people thoughts that are worth arousing" (course 4). This pedagogy approach may develop learners' literacy and intercultural skills based on activism for inclusion, tolerance, and initiative to promote response to social-cultural diversity and learning achievement.
ReferencesAdler, P. A. & Adler, P. (1994). Observation techniques. In Norman K. Denzin & Yvonna S. Lincoln (Eds.), Handbook of qualitative research (pp.377–392). Thousand Oaks, CA: Sage.
Blaschke, L.M. (2021). The dynamic mix of heutagogy and technology: Preparing learners for lifelong learning. British Journal of Educational Technology, 52, 1629-1645. https://doi.org/10.1111/bjet.13105
Blaschke, L. M. & Hase, S. (2021). So, you want to do heutagogy: principles and practice. In S. Hase & L.M. Blaschke (Eds.), Unleashing the power of learner agency. EdTech Books. https://edtechbooks.org/up/pp
Butler, J., & Milley, P. (2020). Teacher Candidates' Policy Agency to Reframe the Meaning of Citizenship in the Ontario Secondary School Curriculum. Canadian Journal of Education, (4), 1131–1159.
Creswell, P., Cheryl N. author, & Hall, Molly Indexer. (2018). Qualitative inquiry & research design: Choosing among five approaches (Fourth edition). Sage Publications, Inc., Thousand Oaks.
Corbin, J. & Strauss, A. (2008). Basics of qualitative research 3e. Sage.
Glassner, A. & Back, S. (2020). Exploring Heutagogy in Higher Education: Academia meets the Zeitgeist. Springer.
Hase, S., & Kenyon, C. (2000). From andragogy to heutagogy. Ultibase Articles, 5(3), 1–10. https://webarchive.nla.gov.au/awa/20010220130000/http://ultibase.rmit.edu.au/Articles/dec00/hase2.htm.
Hase, S. & Blaschke, L.M. (2021). So, you want to do Heutagogy: Principles and Practice. In: Unleashing the Power of Learner Agency (pp. 13-33). EdTechBooks.org.
Hordvik, M., Fletcher, T., Hauge, A.L., Møller, L. & Engebretsen, B. (2021). Using collaborative self-study and rhizomatics to explore the ongoing nature of becoming teacher educators. Teaching and Teacher Education, 101. 103318.
Kawulich, B. B. (2005). Participant Observation as a Data Collection Method [81 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 6(2), Art. 43, http://nbn-resolving.de/urn:nbn:de:0114-fqs0502430.
LaBoskey, V. K. (2004). The methodology of self-study and its theoretical underpinnings. In J. J. Loughran, M. L. Hamilton, V. L. LaBoskey, & T. Rusell (Eds.), International Handbook of self-study of teaching and teacher education practices (pp. 817-870). Kluwer.
Merriam, S. B. (2009). Qualitative research: A guide to design and implementation. Jossey-Bass.
Moore, R. L. (2020). Developing lifelong learning with heutagogy: Contexts, critiques, and challenges. Distance Education, (3), pp. 381–401. https://doi.org/10.1080/01587919.2020.1766949
Pithouse-Morgan, K. (2022). Self-study in teaching and teacher education: Characteristics and contributions. Teaching and Teacher Education, 1038.
Ratnam, T. (2020). Provocation to Dialog in a Third Space: Helping Teachers Walk Toward Equity Pedagogy. Frontiers in Education 5. https://doi.org/10.3389/feduc.2020.569018
20. Research in Innovative Intercultural Learning Environments
Ignite Talk (20 slides in 5 minutes)
Digital Learning Environments for Transformative Education and Intercultural Learning
Luisa Conti
Friedrich Schiller University, Germany
Presenting Author: Conti, Luisa
In the postdigital globalized society, the digital realm seamlessly intertwines with our analog lives, becoming an integral extension of our lifeworld, reminiscent of Negroponte's foresight that "like air and drinking water, the digital will be noticed just by its absence and not its presence" (Negroponte 1998). This hyperdigitalization has significantly impacted the way we perceive and engage in learning, calling for innovative educational design (Bolten 2024). The need for a transformation in education is also linked to the “time of complexity” (Ceruti, 2018), in which we live. It requires citizens who can constructively cope with various challenges and create new, sustainable cultures. In this historical context, education takes on the imperative role of being transformative, offering an experience that goes beyond imparting knowledge and actively shaping individuals with the capacity to consciously transform the reality they live in (UN 2015).
This ignite talk aims to provide insights into two innovative learning environments that leverage digitalization to promote intercultural dialogic learning, placing learners, their diversity, equal rights, and their lifeworld at the center (Author 2022).
The first format is a multilingual platform fostering peer-learning (buddy system) and experiential learning; the second is a simulation game that brings students from different countries together and promotes their collaboration.
Drawing from previous experiences with this format, I expect the ignite talk to facilitate a sharp focus on how these learning environments successfully transform education into an intercultural, transformative experience. It allows for concise highlighting of the core characteristics that led to their success as well its weaknesses.
Methodology, Methods, Research Instruments or Sources UsedQualitative methodology has been employed to assess the potential of the two learning environments in the two different projects. The main research methods applied are: non-participant observation and interviews for both of them, in the second project we also used content analysis of participants' reflection sheets and conversation analysis of their dialogues. This paper aims to compare the results derived from their evaluations and integrate them into a model for digital transformative education.
Conclusions, Expected Outcomes or FindingsThe primary outcome of this paper is the development of a model that synthesizes the characteristics making digital learning concepts promising for transformative education and intercultural learning. Researchers participating in the ignite talk may request a more in-depth exploration of specific findings related to the various projects that led to the experimentation with these distinct learning environments. I will bring therefore also the specific data emerged in the different projects.
ReferencesAuthor (2022). Inklusion durch Dekonstruktion. Der dialogische Ansatz zur Verwirklichung von Inklusion im pädagogischen Bereich. Habilitationsschrift.
Bolten, J. (2024). Scimification Holistic Competence Scenario Development and the Example of Virtual Intercultural Escape Rooms and Strategy Games. In Author & Fergal Lenehan (eds.): Lifewide Learning in Postdigital Societies. Shedding Light on Emerging Culturalities, 29-56. transcript.
Ceruti, M. (2018). Il tempo della complessità. Raffello Cortina.
Negroponte, N. (1998). Beyond Digital. Wired Columns 6(12), retrieved 10.4.2022 from http://web.media.mit.edu/~nicholas/Wired/WIRED6-12.html.
UN (United Nations General Assembly). A/RES/70/1 - Transforming our world: the 2030 Agenda for Sustainable Development. Resolution adopted by the General Assembly on 25 September 2015.
|
