ChatGPT in ELT-Writing Affordances and Activities
ChatGPT in ELT-Writing Affordances and Activities
ChatGPT in ELT-Writing Affordances and Activities
by Allessandra Elisabeth dos Santos, Larisa Olesova, Cristiane Vicentini, and Luciana C. de
Oliveira
To open this article, ChatGPT, a chatbot developed by OpenAI, introduces itself. This
introduction helps us understand, or at least reflect on, the conflicting opinions and uneasy
feelings provoked by its widespread popularity since its launch in November 2022:
This situation reminds us of how previously launched technologies (e.g., smartphones) also
caused fears and enchantment but later became an integral part of our daily social and
educational practices. In fact, and especially, after the COVID-19 pandemic, our social
interactions became increasingly dependent upon digital technologies; therefore, considering that
educators have been continuously examining the effectiveness of various technologies when
utilized as pedagogical resources, our main focus turns to artificial intelligence (AI) and the
pedagogical affordances of ChatGPT.
About ChatGPT
The acronym GPT stands for generative pre-trained transformer. Pre-trained means that when
ChatGPT is used to generate texts, it does not browse the Internet. Instead, it bases the
generation of text upon its previously trained datasets (Brown et al., 2020). ChatGPT’s training
data included sources up to 2021, so at times it might not be the most adequate tool to develop
content consisting of current events. The second term, transformer, indicates that its architecture
was built using machine learning algorithms (OpenAI, 2023b). Also noteworthy is that ChatGPT
is not a standalone language model because it is a subset of GPT-3, a much more powerful
model, released in 2020 (see the generative AI timeline in Appendix A).
As an optimized version of GPT-3’s capabilities, ChatGPT has impressed us with its abilities to
translate language; generate, complete, and classify text; and carry out conversations in diverse
genres (e.g., articles, poems, short stories, and narratives). As such, the outstanding levels of
proficiency often make it hard to distinguish ChatGPT from human-created content.
Additionally, ChatGPT is capable of creating podcast outlines; composing song lyrics; and
writing blog posts, advertisements, and video scripts. ChatGPT can create entire lesson plans and
give individualized feedback on students’ writing. With this in mind, our main goal is to
highlight some pedagogical affordances of ChatGPT to offer different ways in which it can be
integrated into teaching.
Limitations
Before sharing ideas on the ChatGPT integration into teaching, it is worth mentioning some
limitations. The following limitations have been noted by OpenAI (2023b):
As writers are engaged with writing, their minds are often overflowing with thoughts and
emotions. When these emotions are taken for granted, common issues include having writer’s
block, feelings of anxiety, and lack of confidence. Additionally, when writing tasks are assigned,
students at emerging English language proficiency levels may need additional supports in
language and mechanics, such as punctuation and spelling conventions, as well as content related
to the given topic and genre expectations. Although these needs in writing are common for
everyone, they become even more relevant for second language writers. To address this, research
examining how technologies can support writing has reported on how technology tools can
enable peer editing and be used for sharing learners’ work (Vicentini & de Oliveira, 2018).
Findings demonstrate that technology can enhance student motivation, increase engagement, and
scaffold writing.
There have been concerns about the use of ChatGPT for academic writing, particularly regarding
cheating, breach of intellectual property, and creativity. Although these concerns are valid, we
should also consider the numerous affordances of ChatGPT. A major benefit of ChatGPT is that
it can be utilized as a mentor text, which is a great source for scaffolding writing instruction. For
example:
Prompts and Paraphrasing: Instructors can show students how AI can be used as a
resource for writing prompts and paraphrasing. Teachers can have students place
different classmates’ entries (or examples from textbooks) into ChatGPT and have it
Teacher Professional Development: Boa Sorte et al. (2021) described how ChatGPT’s
algorithm can be applied and implemented into writing practice, using it for teacher
reflection, planning, and preparation.
Based on these examples, and considering that scaffolding strategies are known to enhance
writing development and should be incorporated into writing instruction, we make a case for the
integration of ChatGPT as a pedagogical resource due to its numerous affordances. Next, we
offer practical examples of how the tool can successfully enhance writing instruction.
Figure 1. Revised Bloom’s Taxonomy. (The Center for Teaching, Vanderbilt University, 2016,
CC BY 2.0)
Considering that remember and understand require learners to retrieve prior knowledge, you can
ask learners to list or summarize, recognize or classify, recall or clarify, and identify or predict.
ChatGPT can be integrated into these tasks as an ideas generator, used for brainstorming and
helping learners spark further thinking. For example, brainstorm ideas for a podcast script (see
Appendix B for a full lesson plan).
Apply
The apply level of the taxonomy can be used for discussions. Learners can respond to frequently
asked questions or provide advice. For example, ChatGPT can be implemented as an
introductory activity before the discussion so that students can gain awareness, reflect, consider
perceptions, talk, and then discuss (see Appendix B).
Analyze
The analyze level helps learners break materials into foundational parts and determine how those
parts relate to one another. ChatGPT can support achieving this challenging level. Watkins
(2022), among other activities, outlines how learners can use AI to reflect and improve by
Mollick (2023), in his blog “The Practical Guide to Using AI to Do Stuff,” shares how students
can just write anything and then ask ChatGPT how to make their own writing better, or how they
can ask the chatbot to suggest a header or opening paragraph to combat writer’s block.
Evaluation
Finally, ChatGPT can help learners achieve the evaluation level of Bloom’s taxonomy; this is
one of its highest levels and requires learners to make judgements based on criteria and
standards. When using ChatGPT, instructors can create a prompt competition to promote critical
thinking by asking students to develop criteria collaboratively, create prompts individually, and
then pose the questions to ChatGPT and use their predefined criteria to evaluate the responses,
rating and ranking the responses to determine who wrote the best prompts (Watkins, 2022).
Conclusion
ChatGPT has brought numerous and relatively vital calls for reflection, discussion, and change
among educators. Changes such as this can give rise to fear of the unknown and lack of
confidence, and it can leave educators feeling as if there is the constant need for adaptation;
nonetheless, ChatGPT has come to stay. Like previous technologies (the internet or smartphones)
This article demonstrates how ChatGPT can be integrated into education, specifically into
writing classes. Ideas shared in this article can help teachers, teacher educators, schools, and
higher education institutions to consider creative ways to start using ChatGPT to develop
students’ higher levels of critical thinking.
References
Anderson, L. W., Krathwohl, D. R., Airasian, P. W., Cruikshank, K. A., Mayer, R. E. Pintrich, P.
R., Raths, J., & Wittrock, M. C. A. (Eds.). (2001). A taxonomy for learning, teaching, and
assessing: A revision of Bloom’s taxonomy of educational objectives. Addison Wesley Longman.
Boa Sorte, P., de Freitas Farias, M. A., dos Santos, A. E., do Carmo Andrade Santos, J., & dos
Santos Rodrigues Dias, J. S. (2021). Artificial intelligence in academic writing: What is in store
with the GPT-3 algorithm? Revista EntreLinguas, 7(00), 1–22.
https://doi.org/10.29051/el.v7i00.15352
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A.,
Shyam, P., Sastry, G., Askell, Agarwal, S., Herbert-Voss, A., Krueger, G., Henighan, T., Child,
R., Ramesh, A., Ziegler, D. M., Wu, J., Winter, C., … Amodei, A. (2020, May 28). Language
models are few-shot learners. https://arxiv.org/pdf/2005.14165v1.pdf
Vicentini, C., & de Oliveira, L. C. (2018). Using technology tools in writing instruction. NYS
TESOL Journal, 5(2), 44–51.
http://journal.nystesol.org/july2018/6VicentinideOliveira%28CGFP%29.pdf
Vicentini, C., de Oliveira, L. C., & Gui, J. (2022). Integrating technology into genre-based
writing instruction for multilingual learners. GATESOL Journal, 32(2), 27–42.
https://doi.org/10.52242/gatesol.167
_______________________________
Larisa Olesova, PhD, is a clinical assistant professor in the School of Teaching and Learning at
the University of Florida. Her research focuses on distance education, specifically asynchronous
online learning environments. Other areas of research and practice include aspects of online
presence, the community of inquiry, instructional strategies, and best practices in online
teaching.
Luciana C. de Oliveira, PhD, is associate dean for academic affairs and graduate studies and a
professor in the School of Education at Virginia Commonwealth University. Her research
focuses on issues related to teaching multilingual learners in K–12, including the role of
language in learning the content areas and teacher education. She is a past president of TESOL
International Association.