Irish Journal of Technology Enhanced Learning, Vol 7, Issue 2
Special Issue: The Games People Play: Exploring Technology Enhanced
Learning Scholarship & Generative Artificial Intelligence
https://doi.org/10.22554/ijtel.v7i2.151
Aodán Farrelly*
University of British Columbia
This paper offers a reflective narrative detailing the utilisation of generative artificial intelligence (GenAI), with a focus on ChatGPT 3.5, HIX.AI, and Perplexity, in the construction of a concise report on Hybrid and HyFlex teaching models. The writing approach encompassed the formulation of prompts directed at GenAI, tackling issues related to naming conventions, student implications, and strategies for effective teaching in Hybrid and HyFlex settings. Following the presentation of the report produced by GenAI, the paper culminates in a critical reflection on the overall writing process involving GenAI and considers the possible implications. The overall sense from this exercise is that free versions of GenAI can produce a reasonably competent and articulate piece of work but only with time, patience and an existing prior knowledge base that enables the ‘author’ to be able to evaluate and interact with the AI output.
While I was aware of AI in its various forms for a number of years, for me and many others, I suspect, the arrival of ChatGPT in late 2022 proved to be a genuine game changer in terms of raising my awareness and subsequent level of use. I was aware that there were accusations of bias in terms of the algorithms and the limitations in terms of incomplete information, most notably that it did not have access to real-time information that stopped in September 2021. Prior to this exercise, I had only used ChatGPT-3 and latterly version 3.5.
In terms of selecting the topic for this paper, I was drawn to the subject of Hybrid and HyFlex teaching and learning both in terms of issues around (1) the clarity of understanding of the two terms and (2) the pedagogical and technical considerations for teaching in a Hybrid mode. The primary reason for this choice is that this was the subject matter for my research project when undertaking an MA in E-Learning Design and Development in 2021-2022 and drawing on my experience of working as a digital learning support officer in a large university. As such, I felt confident that I had the knowledge to be able to evaluate the efficacy, accuracy and appropriateness of the AI outputs.
Based on the insights from my work experience and my previous research I was aware that there appears to be a degree of confusion as to what exactly constitutes Hybrid, blended and HyFlex learning. This search for clarity is not simply an exercise in semantics about naming conventions. Howell (2020 p.63) concludes that “names matter, particularly in emerging or evolving teaching contexts where teachers, students, and researchers are in the process of defining and refining the multiple subjectivities actors may occupy”. Saichaie’s (2020) comprehensive analysis of the naming conventions associated with technology-enhanced learning found that the terms blended, flipped and Hybrid were in many instances used interchangeably. This has resulted in varying degrees of confusion and differential expectations on the part of the teaching staff, institutional management, and students (Raes, 2020). What characterises Hybrid teaching is that the learning space consists of two synchronous modes with students able to attend either on-campus or online in a synchronous environment (Irvine, 2020; Lamb et al. 2021). Because of my familiarity with the topic, I was interested to see if AI had a grasp on the nuances of the debate and the ensuing implications for digital education practice.
Following on from my own master’s research, I wanted AI to write a short report about naming conventions and the issues associated with confusion between Hybrid, blended and HyFlex modes followed by a practice-based focus second half that offers advice to teaching staff using a Hybrid and/or HyFlex approach. However, rather than set the AI one overarching prompt whereby I was asking it to write a 2,000-3,000-word complete paper, I used GenAI by posing a series of questions ranging from issues about naming conventions to asking for advice on how to lecture in a Hybrid/HyFlex mode. Thus, the prompts used were:
As can be seen from the list of prompts, numbers five to eight also included a request for references. As presented subsequently in the short report, I have indicated where a specific reference was supplied by the AI with a number from 1 to 6. Otherwise, there was a series of general references supplied but not attributed to a specific point, rather it was an attribution indication of where the general points had been drawn from.
It is important to acknowledge that the paper is not intended or presented as an exhaustive or systematic evaluation of different GenAI platforms, rather, this paper sets out to report and reflect on the issues encountered in the writing and research process. Perhaps not unsurprisingly, I started with, and primarily used ChatGPT 3.5 during the process. While being mindful of the reported limitations and bias of ChatGPT (Rawas, 2023); due to its sheer pervasiveness, it was the primary tool used in this project. However, once the initial set of responses was generated, the prompts were copied and pasted into two other generative AI tools: Perplexity (https://www.perplexity.ai/) and HIX.AI (https://hix.ai/chat).
The subsequent short report article was ‘stitched’ together in response to the different prompts and from the three GenAI platforms indicated. In order to enhance the transparency of the process I have provided an indicator in terms of the platform used to generate the text and what prompt it was written in response to using the letters P for Perplexity, C for ChatGPT and H for HIX; hence prompt number 3 from ChaptGPT will be indicated at the end of the relevant section as [#3C] and so on. Finally, when the main body of the report was complete, I submitted the text to ChatGPT and asked it to first write an introduction and then write a conclusion. In both instances, these were the first iterations used for the report. Finally, when I completed the short report, I again submitted it to ChatGPT and asked for a list of suggested titles, from which I selected the final title. The only author-generated text in the report is the headings that I provided in order to help with structure and presentation.
Hybrid teaching and HyFlex teaching are two prominent instructional approaches that merge in-person and online learning methods. These methodologies, though bearing some similarities, also possess distinct differences. In this article, we delve into the core distinctions between Hybrid and HyFlex teaching models, explore the implications of their usage in educational contexts, and provide strategies for instructors to ensure equivalence of experience for both on-campus and remote students in a Hybrid teaching environment. The ever-evolving landscape of education demands a comprehensive understanding of these approaches to meet the diverse needs of today's students.
Hybrid teaching and HyFlex teaching are both instructional approaches that combine in-person and online learning. While there are similarities between the two, they also have distinct differences. Here are the key differences between Hybrid and HyFlex teaching:
Definition: Hybrid teaching refers to a model where students participate in both in-person and online activities at the same time. It typically involves dividing students into two groups and rotating them between physical classrooms and online learning platforms. On the other hand, HyFlex teaching stands for Hybrid-Flexible, and it offers students the flexibility to choose between attending classes in person or participating remotely, using synchronous or asynchronous learning options.
Flexibility: Hybrid teaching has preset schedules for students to alternate between physical and virtual classrooms, with a predetermined arrangement for in-person and online activities. In contrast, HyFlex teaching prioritizes student choice by offering them the flexibility to decide how they want to participate in each class session. This could include attending in person, joining remotely via video conference, or accessing class materials asynchronously.
Technology Usage: Both models rely on technology integration, but they differ in emphasis. Hybrid teaching often adopts video conferencing tools to connect remote learners with the class. However, the primary focus is on delivering synchronous instruction and maintaining simultaneous learning experiences for all students. HyFlex teaching, on the other hand, places more emphasis on leveraging technology to deliver content and engage students asynchronously, recognizing that students may have different learning preferences and time constraints.
Pedagogical Approach: In Hybrid teaching, the pedagogical approach tends to be more synchronized, with both in-person and online methods following a similar structure and pacing. The aim is to ensure consistency in content delivery and to create a cohesive learning experience for all students, regardless of their mode of participation. In HyFlex teaching, the pedagogical approach may vary depending on how students choose to participate. The instructor must design activities and assignments that allow for individual paths of exploration, fostering deeper engagement with the material [#1C].
Both approaches have similarities and differences. Here are some of the similarities and differences between Hybrid and HyFlex teaching:
Similarities
Differences
In summary, while Hybrid teaching requires students to alternate between in-person and online learning, HyFlex teaching offers students the flexibility to choose their own learning mode for each class session. The focus of HyFlex teaching is to provide student choice, while Hybrid teaching aims for a more synchronized learning experience [#1H].
It is important to note that these terms are not always used consistently, and there is some overlap between them. However, understanding the differences between these terms can help educators and students choose the approach that best suits their needs. [2P]. Here's a discussion of why this happens and the distinctions between these terms: Lack of Standardized Definitions: One of the main reasons for the confusion is the absence of standardized definitions for these terms in the education field. Different institutions, educators, and researchers may interpret and use these terms differently. Evolution of Terminology: Over time, the meanings of these terms have evolved. What was once considered "blended" or "Hybrid" in the past might now align more closely with the concept of HyFlex due to advancements in technology and changing teaching methodologies. Contextual Variations: The interpretation of these terms can vary depending on the context. For example, a school or university might have its own specific definition for what constitutes a "blended" or "Hybrid" course, and this definition may differ from that of another institution. HyFlex as a Hybrid Evolution: HyFlex is a relatively newer concept compared to Hybrid and blended learning. It emerged as a response to the need for more flexibility in education, especially during situations like the COVID-19 pandemic. As a result, some educators and institutions may use "HyFlex" to describe courses that others would categorize as "Hybrid" or "blended". To mitigate confusion, it's essential for educational institutions and instructors to provide clear definitions and guidelines when using these terms. This helps ensure that students, faculty, and administrators have a common understanding of the instructional approach being employed. Additionally, as these terms continue to evolve, the education community should work toward creating more standardized definitions to reduce ambiguity and promote effective teaching and learning practices [#2C].
The lack of clarity in institutional naming conventions for the use of Hybrid and HyFlex teaching models in higher education can have several implications for students:
In summary, lack of clarity in institutional naming conventions for Hybrid and HyFlex courses can lead to confusion, inconsistent learning experiences, and other challenges for students. It's crucial for institutions to address this issue by providing clear information and ensuring that both faculty and students understand the meaning and expectations associated with these terms. These measures can help ensure that students make informed choices, experience equitable learning opportunities, and have a positive overall educational experience [#4C].
To mitigate these implications, institutions should prioritize clear and consistent communication regarding course formats and expectations. This may include: Establishing standardized definitions and guidelines for the use of terms like "Hybrid" and "HyFlex." Providing detailed course descriptions, syllabi, and registration information. Clearly communicating any technology requirements for online participation. Offering training and support for faculty to implement teaching models effectively and consistently. Creating feedback mechanisms for students to express concerns and questions about course formats. These measures can help ensure that students make informed choices, experience equitable learning opportunities, and have a positive overall educational experience. For specific sources addressing these implications, you may refer to articles and publications on higher education practices and policies related to course design and communication #[4C].
Teaching in a Hybrid or HyFlex learning environment, where you simultaneously engage with in-person and remote students, can be a complex task. Here's some advice for a lecturer who will be required to teach in such settings:
Ensuring equivalence of experience between on-campus and remote students in a Hybrid teaching environment can be challenging, but there are several strategies that can help lecturers create a flexible, accessible, and equitable learning environment for all students. Here are some ways to ensure equivalence of experience [8 P]:
These strategies will help you maintain a balanced and equitable learning experience for both on-campus and remote students in a Hybrid teaching environment. Regular assessment of the effectiveness of these strategies and feedback from students can further guide adjustments and improvements to ensure equivalence. By incorporating these strategies, instructors can create a more inclusive and engaging learning environment for remote students in a Hybrid or HyFlex teaching setting. [#8C].
In conclusion, the distinctions between Hybrid and HyFlex teaching models are crucial to understand, as they have a significant impact on the learning experiences of students in higher education. Hybrid teaching involves a predetermined combination of in-person and online activities, while HyFlex teaching empowers students to choose their mode of participation for each class session. While the terms are sometimes used interchangeably, this lack of consistency can lead to confusion and disparities in learning experiences.
The implications of unclear naming conventions for students are substantial, including confusion, inconsistent learning experiences, and difficulties in planning their academic journeys. To address these challenges, educational institutions must prioritize clear and consistent communication, standardized definitions, technology readiness, and support mechanisms. These measures will help ensure that students make informed choices, experience equitable learning opportunities, and have a positive overall educational experience.
For instructors teaching in Hybrid or HyFlex environments, adaptability, technology proficiency, and a strong commitment to engaging all students are key. Flexibility in course design and clear communication about expectations and technology requirements are essential for fostering a successful learning experience.
Ultimately, achieving equivalence of experience in a Hybrid teaching environment is an ongoing process that requires careful planning, effective technology utilization, and a commitment to promoting fairness and inclusivity in higher education. As the landscape of education continues to evolve, addressing these challenges will remain a priority for educators, institutions, and policymakers alike.
The reference list for the sources supplied by ChatGPT 3.5 and HIX.ai are reproduced as they were presented. The references in Perplexity were simply indicated as numbers in the text and web links at the bottom of the page. I have reproduced the HIX.ai references in APA style for readability.
The University of Arizona. (n.d.). What is HyFlex? [Website]. https://teachingonline.arizona.edu/HyFlex
Educause Learning Initiative. (2010). 7 Things You Should Know About the HyFlex Course Model [PDF]. https://library.educause.edu/resources/2010/10/7-things-you-should-know-about-the-HyFlex-course-mode l
Bates, T. (2019). Definition of Hybrid Learning [Webpage]. https://www.tonybates.ca/2019/08/30/what-do-we-mean-by-Hybrid-learning/
Beatty, B.J. (2019). Hybrid-Flexible Course Design: Implementing student-directed Hybrid classes. EdTech Books. Link
Graham, C., Dziuban, C., & Moskal, P. (2013). Course Design, Time Investment, and Student and Instructor Perceptions in Blended and Online Learning Environments. Journal of Online Learning and Teaching, 9(3), 349-362.
Over a two-day period using different prompts and iterations I generated a total of 11,400 words to eventually produce the short report of 2,600 words long with another 300 in the reference list. This generation phase was followed by the subsequent selection and verification phases, which took place over a further four-day period. Between the two phases, I would guesstimate that the time involved in producing the short report was approximately 8-10 hours.
Because of the sequential way the prompts were used and because free versions of the platforms have less functionality, the report was, as previously stated, ‘stitched’ together. There was no author-generated text inserted to bridge between sections, consequently, selections of responses were made on the basis of how accurate the information in each section was, and how the quality of the writing could be used to best create a sense of flow between sections. Because of the staggered nature of writing with a series of prompts the outputs produced were perhaps a little more staccato in terms of most bullet-point narrative, frequently finished by the stock phrase ‘in summary’. However, with careful and judicious selection and sequencing of the outputs, the final report is (given the required word limit) I believe, an acceptably comprehensive paper that provides a reasonably well-nuanced understanding of the topic and the associated teaching and learning issues that would provide the novice in the world of Hybrid and HyFlex teaching a useful starting point.
One of the principal criticisms about GenAI is its reported tendency to what is often referred to as hallucinations, in other words, the tendency to produce “plausible, but false information in its output (such as fake references) and so being able to evaluate the output for its quality is a key capability in making use of AI tools” (Hillier, 2023). Therefore, in making my determinations I paid particular attention to the accuracy and appropriateness of the references supplied. However, while my previous experience of ChatGPT in the early part of 2023 did highlight examples of hallucinations, this round of searches produced very different results with the three specific references supplied (see the previously listed ChatGPT references) all proving to be factually correct and appropriate. In response to prompt #5, ChatGPT produced an equally noncommittal response, stating that “these sources provide general information about Hybrid and HyFlex teaching models. Keep in mind that the specific implementation and terminology can vary among institutions and instructors”.
At the time that I used the HIX.AI platform, I was not initially aware that the text generator is essentially ChatGPT 3.5; therefore, the same issue with the provision of general references similar to the direct use of ChatGPT 3.5 was evident. In the case of the Perplexity outputs, specific references were indicated at the end of each paragraph as indicated in the short report. As such, there was greater transparency in terms of attribution in comparison to the other two platforms used in this exercise. The appropriateness of the seven references supplied by Perplexity was appropriate in terms of the correctness of the information supplied and the provenance of the authors.
As I said in section one, aside from ChatGPT I had not used GenAI to any great extent. In fact, the extent of my knowledge was illustrated when halfway through the writing phases I realised that HIX.ai has a limit in terms of free outputs, after that limit has been reached, a subscription service is required. Therein lies one of the issues regarding the different versions of GenAI, namely the digital divide now has another layer; whether one has the resources to be able to pay for subscription versions of GenAI platforms with their far higher levels of capability and functionality. For example, in arguing that AI could widen the digital divide, McKean (2023) argues that “Initial estimates by the experts at Jisc’s National Centre for AI (NCAI) found that if a student were to subscribe to a full suite of popular generative AI tools and education plug-ins, it could cost them around £1,000 a year, pushing generative AI out of reach for many”. Even my brief exposure to the limitations of HIX.ai illustrated very quickly to me the potential shortcomings of relying solely on free versions.
Given the rapid pace of change in one year it is almost impossible to guess where GenAI will take us and what the possible implications are for higher education. Aside from the aforementioned ‘digital divide’; there appears to be another divide at work on this topic; namely, there are those who fear the ‘traditional’ written assessments have had their day and those who argue that teaching, learning and assessment could be largely redundant. On the other hand, there are those who welcome the opportunities presented by GenAI to be able to reimagine how higher education teaches and assesses. As someone who has been deeply involved in digital education for several years, I tend to be in the latter camp. However, as this exercise has demonstrated to me, I was only able to construct the Hybrid/HyFlex paper on the basis of prior knowledge understanding and skills; without that knowledge and informed judgement I am quite sure that I would have had the ability to construct the finished paper. In considering the possible implications for practice, my views and reflections are based on my role in educational technology design and support rather than as a student-facing lecturer. From my perspective, GenAI presents an opportunity for streamlined content creation and enhanced personalised learning experiences, thus reshaping the instructional design process. Whether that reshaping is for better or worse? only time will tell.
Hillier, M. (2023). A proposed AI literacy framework. TECHE. https://teche.mq.edu.au/2023/03/a-proposed-ai-literacy-framework/
Howell, D. C. (2020). "A rose by any other name" - The significance of naming in face-to-face, online, and hybrid teaching. Curriculum and Teaching Dialogue, 22(1-2), 61–75.
Irvine, V. (2020, October 26). The landscape of merging modalities. Educause Review, 55(4), 41-58. https://er.educause.edu/articles/2020/10/the-landscape-of-merging-modalities
Lamb, J., Caravalho, L., Gallagher, M., & Know, J. (2022). The postdigital learning spaces of higher education. Postdigital Science and Education, 4(1), 1-12. https://doi.org/10.1007/s42438-021-00279-9
McKean, P. (2023, September 12). Without intervention, AI could widen the digital divide for students. JISC Blog. https://beta.jisc.ac.uk/blog/without-intervention-ai-could-widen-the-digital-divide-for-students
Raes, A. (2022). Exploring student and teacher experiences in hybrid learning environments: does presence matter? Postdigital Science and Education, 4(1), 138–159. https://doi.org/10.1007/s42438-021-00274-0
Rawas, S. (2023). ChatGPT: Empowering lifelong learning in the digital age of higher education. Education and Information Technologies. https://doi.org/10.1007/s10639-023-12114-8
Saichaie, K. (2020). Blended, flipped, and hybrid learning: Definitions, developments, and directions. New Directions for Teaching and Learning, 2020(164), 95–104. https://doi.org/10.1002/tl.20428