See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/381663266
From AI-Generated Lesson Plans to the Real-Life Classes: Explored by Pre-
Service Teachers
Conference Paper · June 2024
DOI: 10.4995/HEAd24.2024.17060
CITATION READS
1 779
2 authors, including:
Dilara Yılmaz Can
Kocaeli University
22 PUBLICATIONS 50 CITATIONS
SEE PROFILE
All content following this page was uploaded by Ceyda Durmuş on 24 June 2024.
The user has requested enhancement of the downloaded file.
10th International Conference on Higher Education Advances (HEAd’24)
HEAd Universitat Politècnica de València, València, 2024
DOI: https://doi.org/10.4995/HEAd24.2024.17060
From AI-Generated Lesson Plans to the Real-Life Classes:
Explored by Pre-Service Teachers
Dilara Yılmaz Can1 , Ceyda Durmuş2
1Kocaeli University, Turkey, 2Republic of Türkiye Ministry of National Education, Turkey
How to cite: Yilmaz Can, D. & Durmus, C. 2024. From AI-Generated Lesson Plans to the Real-Life
Classes: Explored by Pre-Service Teachers. In: 10th International Conference on Higher Education
Advances (HEAd’24). Valencia, 18-21 June 2024. https://doi.org/10.4995/HEAd24.2024.17060
Abstract
ChatGPT is a powerful Artificial Intelligence (AI) technology that has the potential to
revolutionize the way we study in education. It can be used in a variety of ways and we,
on the other hand, focused on senior pre-service teachers’ designing a mathematics
lesson plan to implement in primary schools. Four voluntary participants attended, as
half of which designed a mathematical task by asking AI while the rest used traditional
methods in planning the task. In the next breath, participants implemented their lesson
plans in a public school and collected the data relating to tasks they used. At the end, we
compare the outcomes concerning the lesson plan and implementation in schools, and
determine the strengths and weaknesses of each approach. This research offers valuable
insights into using ChatGPT in teacher education to be used in practice and considering
the teachers’ task designing role in classrooms with and without ChatGPT.
Keywords: Artificial intelligence; ChatGPT, teacher education, pre-service teachers,
mathematics lesson plan
1. Introduction
Once, it was education leading the way in paradigm shifts. Now, propelled by research and
development activities— in other words, by riding the wave of artificial intelligence's [AI]
meteoric rise—it has become inevitable to compel the four horses of education to pursue
novelty. AI-generated tools elegantly dance onto the stage. Generative AI is stirring up quite the
excitement, thanks to using the magical blend of natural language processing and extensive
language models. Because AI has this superpower to grasp and spit out human language which
is known to be the pinnacle of sophistication in system development, to spot patterns; the
trained-up versions of this AI can shuffle through language constructs (Bozkurt, 2023). While
it took humans approximately six million years to evolve, we now possess a polyglot and
talkative genius at our fingertips, capable, if trained proficiently, of swiftly solving previously
unsolved problems in mathematics.
This work is licensed under a Creative Commons License CC BY-NC-SA 4.0
Editorial Universitat Politècnica de València 44
From AI-Generated Lesson Plans to the Real-Life Classes
Generating quite the mojo with its content-creating prowess, generative AI has sparked a flurry
of scholarly debates, prompting researchers to dive into its potential perks and pitfalls (Johinke
et al., 2023). Authorities in the realms of education and innovation proclaim that given the
burgeoning presence of generative AI tools in the lives of young individuals, a strategic
endeavor is imperative to blend them seamlessly into educational frameworks. Knowing that AI
is used in numerous fields (Kasneci et al., 2023), we attempt to answer this question in the
educational scope. This research seeks to explore the effects of incorporating AI-generated
lesson plans into the teacher education program, center upon the viewpoints and encounters of
future educators.
One of the AI tools that can be employed for educational purposes is ChatGPT AI language
model developed by OpenAI. It's designed to understand and generate human-like text based on
the input it receives. Using a vast amount of data and machine learning techniques, ChatGPT
can engage in conversations, answer questions, generate text in various styles, and assist with a
wide range of tasks, from providing information to creative writing. What makes ChatGPT
unique is its ability to produce high-quality text for broad input, constant learning capabilities,
and partly free usage. ChatGPT showcases its capacity to bolster student writing through
multiple avenues: offering outlines, scaffolding knowledge, generating summaries,
paraphrasing content, providing feedback on diverse submissions, and even facilitating practice
quizzes on specified subjects (Halaweh, 2023).
ChatGPT offers a spectrum of support in educational settings, from automated feedback,
discussion starters, and grammar/spelling corrections to crafting syllabi, lesson plans, and
developing curricula (Cope et al., 2021; Lin, 2022). Numerous recent studies have proposed that
the utilization of ChatGPT within educational contexts presents opportunities for enhanced
teaching methodologies and improved learning experiences (Adiguzel et al., 2023; Cooper,
2023; Kasneci et al., 2023). When it comes to core subjects, the integration of lesson plan
preparation within teacher education programs, with a particular focus on mathematics may
affect the classroom setting at the schools. Hays et al., (2023) found that the majority of teachers
do not use any AI tools and ChatGPT either in their lessons. And the minority use it for the
lesson plans/substitute plans, letters of recommendation and asking how to teach with it.The AI
platform, by identifying a student's areas of struggle in mathematics, can dynamically adapt the
lesson (Graham, 2023). Spot on the present research, Hong (2023) and Phillips (2023) have
discussed aspects of lesson planning. However, no research articles that focus on GPT-generated
lesson plans in mathematics and also carry out the plans in primary schools to see if it works in
real practice. This study aims to investigate the impact of integrating AI-generated lesson plans
into the teacher education curriculum, focusing on the experiences and perspectives of pre-
service teachers. Based on this, we specify the research question:
1) What similarities and differences exist between Chat-GPT and the traditional
mathematics lesson plans created by pre-service teachers?
45
From AI-Generated Lesson Plans to the Real-Life Classes
2) What implications arise from varying learning plan applications for pre-service
teachers and primary students?
2. Methodology
This research is comparative case study which is a research approach to formulate or assess
generalizations that extend across multiple cases. Yin (1984) states that case studies in
qualitative research assist in researching complex phenomena (cases) or those that we do not
know much about. The cases in this research are designing a mathematical task by the aid of
ChatGPT and without the ChatGPT support. Research trajectory is explained below.
AI-generated lesson planning workshop and creating the lesson plan: We conducted a two hour-
workshop where pre-service teachers are introduced to ChatGPT and guided in using the AI
model to generate lesson plans. Three pre-service teachers participated in this training while the
other three participants who will conduct the lesson with traditional methods were out in that
part of the research. Pre-service teachers created the lesson plan in accordance with their topic
in accordance with the 5E model from constructivist approach and right after the workshop and
sent it to the researchers. We specify to study “fractions” as a mathematical theme so that all
participants are coherent with the implication of the theme. We employed three fraction
concepts for every two of the pre-service teachers (one uses AI, one uses a traditional method)
to design a lesson: i) Instruction of unit fractions’ magnitude, ii) Instruction of proper fractions’
magnitude, iii) fraction comparison.
Implementation in school visits and observations: Whole process takes place in the same public
school where the classrooms include 24, 4th grade students on average. Each attainment was
implemented in the same classroom subject to two different lesson plans. In other words, one
attainment was assigned to a class and the attainments were implemented as two practices in 3
different classes. Researchers gather observations and feedback from both pre-service teachers.
As a sum, the observation form is filled by researchers and by pre-service teachers regardless
of implementer or observer.
Comparative analysis: Texts generated by ChatGPT and tradition-based lesson plans were
utilized for document analysis. The researchers ensured that, through credibility, the data were
collected from the official ChatGPT site. Participants’ reflections, their notes, and researchers’
observation notes were assessed, and all data were cross-checked for better implications. All
data were analyzed collectively by the researchers, and decisions were reached by consensus.
3. Findings
3.1. Assessment of lesson plans
The lesson plans with a traditional lens and with the support of ChatGPT were analyzed
according to the six criteria on Table 1.
46
From AI-Generated Lesson Plans to the Real-Life Classes
Table 1: Assessment Results of Participants’ Lesson Plans
Method & Techniques Originality
AI-generated Traditional AI-generated Traditional
-Question & answer -Question & -Highly original -Mostly conventional
-Narrative answer tasks and activities
-Discussion -Demonstration -Depends on the planner
-Group working
Functionality Suitability
AI-generated Traditional AI-generated Traditional
-Supporting various strategies -Emphasis on Inappropriatenes -More flexible to arrange
-Promoting peer teaching and visual s to student level the plan according to the
sharing representation and grade level suitability
-Giving otantic example and modeling -Redirection to
-Offering accessible materials -Detailed arithmetic
-Rule-based approach explanations processing
According to Table 1, ChatGPT offers group working and discussion regarding methods and
techniques. Question and answer methods are standard for both tools. Oppositely, ChatGPT
generates more original ideas. Besides, traditional lesson plans offer more convenient content
and require less inference. As for functionality, supporting various strategies, promoting peer
teaching and sharing, and giving real-life examples are considered positive features of AI-
generated tools. Emphasis on visual representation and modeling and detailed explanations are
positive sides of traditional lesson plans. It can be accepted that quite similar lesson plans were
created by offering easily accessible materials such as papers, cardboard, and visual models.
3.2. Assessment of Pre-Service Teachers’ Implications
The lessons with a traditional lens and with the support of ChatGPT were analyzed according
to the strengths and weaknesses based on the observations of researchers and participants
themselves.
According to Table 2, while the ChatGPT lesson plan stands out for providing creative and
varied ideas and saving time, the weaknesses observed by both the researchers and the
participants were the lack of depth/superficiality in the subject matter, the predominance of the
narrative technique, and the plans that were not suitable for the grade level and the level of the
students. The timing of the plan stages is a deficiency because it gives very short time for
essential tasks, for example, while it gives much longer time for the reminder stage. As a
weakness, it was also noteworthy that it did not provide any technological support and did not
47
From AI-Generated Lesson Plans to the Real-Life Classes
Table 2: Assessment of Participants’ Implications with AI-generated Lesson Plan
Implications with AI-generated Lesson Plan
Strengths Weaknesses
To implementers To observers To implementers To observers
-The -Proposing -Superficiality -Superficiality
predominance of number line -Dominance of -Dominance of narrative technique
the question & activities narrative techniq -Inappropriate for grade level
answer technique -Providing -Inappropriate -Not offering technology support
-Saving time examples of for grade level -Proposing number line activities
-Offer different effective -Inappropriate -Inadequate contextual problem
perspectives and reminders for student level -Professional knowledge deficiency
diversity -Suggestions on -Lack of joy -Direction to arithmetic process
material -Timing
Implications with Traditional Lesson Plan
Strengths Weaknesses
To implementers To observers To implementers To observers
-Leaving free -Leaving free -Offering single -Adherence to the textbook
space for the space for teacher point of view
teacher -Classroom
-Supporting management
authentic learning effectiveness
suggest any contextual problems. Providing compelling reminder examples was interpreted as
a strength because the comprehensibility of the topic needs to provide reminders about fraction
concepts, as the last topic was seen about a year ago. Finally, the observers noticed a deficiency
in participants' professional knowledge and redirection to arithmetic processing.
While the prominent strengths of traditional lesson plans are leaving free space for the
practitioner, flexibility, and supporting authentic learning, the weaknesses are offering a single
point of view depending on the decision of the implementer and adherence to the course
materials. As a final step, participants were asked: What are the differences between the lesson
plan and implementation with the other method?
PT1: "In the traditional plan, the pre-service teacher prepared a more flexible plan. When I
stayed connected to the chatGPT, I felt as if I was making a presentation rather than teaching,
so there was a one-sided communication."
48
From AI-Generated Lesson Plans to the Real-Life Classes
PT4: "Tools such as ChatGPT may harm the teacher's productivity and creativity in the long
run. But it saves a lot of time when preparing lesson plans. I think traditional and AI generated
can be used together."
PT3: "A plan designed for 1 hour requires at least 3 hours. Timing is a big problem"
PT2: "Despite all the weaknesses, if I didn't know beforehand, I wouldn't have realized that it
was prepared with AI."
4. Discussion and Conclusion
In this research, while asking about the differences between traditional and ChatGPT-supported
lesson plans, we wanted to answer the question, "Which one provides more effective teaching?"
It is not easy to give a simple answer to this. Traditional does not always mean unuseful and
old, but it does not mean that we should always stick to it. This proposal not only explores the
practical implications of AI in lesson planning but also delves into the perceptions and
experiences of those directly involved in the educational process—student teachers. It adds a
valuable dimension to the ongoing discourse on AI in education and teacher training.
The suggestion of visuals and modeling as a common strength of ChatGPT and the traditional
lesson plan is essential because visual representations in mathematics teaching make the subject
concrete and comprehensible (Goldin, 2008). Contrary to traditional lesson plans, ChatGPT
presents various viewpoints and diversity, offering imaginative and diverse suggestions.
Delving deeper into the concept of creativity is valuable, as aspiring educators highlighted its
role in generating innovative lesson plans and fostering their professional expertise. Compatible
with this result, van den Berg & du Plessis (2023) assesed that ChatGPT can serve as a potent
resource in honing educators' abilities for critical thinking. This tool empowers them to assess,
modify, and enhance their teaching methodologies, thereby catering more effectively to their
students' requirements. Considering the activities generated by ChatGPT can certainly expand
teachers' viewpoints and nurture creativity by introducing various approaches to teaching
fractions. This exposure inspires educators to explore inventive methods which can captivate
learners in dynamic learning endeavors. Nonetheless, educators may face challenges, as specific
curricula necessitate adherence to strict guidelines and designated textbooks. Therefore, while
embracing creative strategies, teachers must tread carefully within these constraints.
Along with the positive sides of AI-generated lesson plans, considerable weaknesses exist. The
prominent situation is that plans are only partially suitable for the grade level and the level of
the students. It offers different and effective content, but the implementers should reconsider its
offer before using it. The participants stated that chatGPT had one-way communication and
were in a situation where the way of presentation was active in the implementation process,
contrary to the interactive environment required by constructivism. The situations that emerged
in practice due to the limited flexibility and the predominance of narrative support these pre-
service teachers' views. According to some participants, ChatGPT may harm the teacher's
49
From AI-Generated Lesson Plans to the Real-Life Classes
productivity and creativity in the long run. While ChatGPT excels in offering precise and
pertinent information, it lacks the capacity for emotional support and the essential human
interaction necessary in teaching (Javaid, et al.,2023). We did not observe such a result in this
study, probably because it required a longer data collection process. As a result, AI-driven
technologies should be used in teacher education to make teaching and learning practices more
effective and efficient. However, teachers should reassess the plans or instructions in ChatGPT
to make it more useful and suitable. The results show that candidates do not want to turn their
backs on Chat but want to use it in a way that allows them to intervene. ChatGPT should be
seen as tools to enhance and supplement teachers' work but not replace the teachers.
As a remarkable result, the AI's suggestion of number line activities is both a weakness and a
strength. It is a strength because the number line is an essential and vital tool for understanding
many subjects in mathematics (Kieran, 1988; Siegler et al., 2011). It is a weakness because
using the number line is a problematic tool for preservice teachers and students (Tunç-Pekkan,
2015). This study interpreted it as students' lack of professional knowledge. As another example
of a negative situation, the contextual problem statement of ChatGPT was only proposed in the
lesson on finding the fractional equivalent of a quantity. The lack of problems with unit fractions
and fraction comparison tasks can make it difficult for students to understand fraction magnitude
and force them into rote learning.
Time and duration are also prominent and can be interpreted differently. ChatGPT is a very
effective and powerful feature at the point of plan preparation and saves much time in the lesson
plan preparation phase. However, it appears to be a limitation and weakness in the
implementation phase. It requires intervention because it gives too many things at the same time
in the same lesson plan. As an echo of this result, according to Duha (2023), the lesson plan in
ChatGPT is at a rudimentary level and requires significant editing. Saving time for the
practitioner can be very useful for a teacher at the beginning of his/her career. However, in
almost all of the lesson plan examples requested from Chat, it was observed that it used the same
way to determine the duration of the lesson sections. Designing lesson plans tailored to specific
subjects and concepts is essential. Consequently, it is anticipated that standardized examples of
the same plan could pose challenges in implementation.
This study has the potential to illuminate the effectiveness and implications of integrating such
tools for prospective teachers. Maximizing the educational benefits of these models requires a
careful approach, including a critical examination of their limitations and potential biases.
Recognizing that these tools support teaching and learning is crucial, emphasizing the
importance of a balanced evaluation to leverage their advantages (Duha, 2023) effectively.
According to Kim & Adlof (2023), we should view ChatGPT as a tool to enhance the learning
process in education rather than as an end in itself for generating artifacts. Despite the various
concerns surrounding using and integrating ChatGPT in classroom settings, we believe that
educators will continue to play central role in haping and guiding students' learning experiences.
50
From AI-Generated Lesson Plans to the Real-Life Classes
References
Adiguzel, T., Kaya, M. H., & Cansu, F. K. (2023). Revolutionizing education with AI:
Exploring the transformative potential of Chat-GPT. Contemporary Educational
Technology, 15(3), 429.
Bozkurt, A. (Ed.). (2023). Unleashing the potential of generative AI, conversational agents and
chatbots in educational praxis: A systematic review and bibliometric analysis of GenAI in
education. Open Praxis, 15(4), 261-270.
Cope, B., Kalantzis, M., & Searsmith, D. (2021). Artificial intelligence for education:
Knowledge and its assessment in AI-enabled learning ecologies. Educational Philosophy &
Theory, 53(12), 1229–1245. https:// doi. org/ 10. 1080/ 00131 857. 2020. 17287 32
Duha, M. S. U. (2023). ChatGPT in Education: An Opportunity or a Challenge for the Future?
TechTrends, 67(3), 402-403.
Hays, L., Jurkowski, O., & Sims, S. K. (2023). ChatGPT in K-12 Education. TechTrends, 1-14.
s https://doi.org/10.1007/s11528-023-00924-z
Hong, W. C. H. (2023). The impact of ChatGPT on foreign language teaching: Opportunities in
education and research. Journal of Educational Technology and Innovation, 5(1).
Javaid, M., Haleem, A., Singh, R. P., Khan, S., & Khan, I. H. (2023). Unlocking the
opportunities through ChatGPT Tool towards ameliorating the education system.
BenchCouncil Transactions on Benchmarks, Standards and Evaluations, 3(2), 100115.
Johinke, R., Cummings, R., & Di Lauro, F. (2023). Reclaiming the technology of higher
education for teaching digital writing in a post-pandemic world. Journal of University
Teaching & Learning Practice, 20(2), 01.
Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., ... & Kasneci,
G. (2023). ChatGPT for good? On opportunities and challenges of large language models
for education. Learning and individual differences, 103, 102274.
Kim, M., & Adlof, L. (2023). Adapting to the Future: ChatGPT as a Means for Supporting
Constructivist Learning Environments. TechTrends, 1-10.
Phillips, R. My Debate with ChatGPT-3. 2023. Available online:
https://robinmarkphillips.com/debate-chatgpt-3/
Sok, S., & Heng, K. (2023). ChatGPT for education and research: A review of benefits and
risks. Available at SSRN 4378735.
van den Berg, G., & du Plessis, E. (2023). ChatGPT and Generative AI: Possibilities for Its
Contribution to Lesson Planning, Critical Thinking and Openness in Teacher Education.
Education Sciences, 13(10), 998.
Yin, R. (1984). Case study research: Design and methods. (3. Basım). California: Sage
Publications.
51
View publication stats
