The Open University Scholarship Exchange
Browse

Final Report: ChatGPT and the distance learner: working with AI to write assignments as the site of teaching and learning

report
posted on 2025-03-26, 16:22 authored by FASSTEST AdminFASSTEST Admin, Zoe DoyeZoe Doye, Sonja RewhornSonja Rewhorn, Edward Wigley

Generative Artificial Intelligence (AI) has made headlines across the world that oscillate between proclaiming existential threats to humanity to ChatGPT being heralded as the beginning of a new era in technology. Whilst the challenges posed by AI to higher education are less dramatic, they have the potential to disrupt the ways in which teaching and learning is delivered to students. Bobula (2024) surveys the nascent scholarship within the field of Generative AI and higher education, identifying common themes including threat to academic integrity, changes to learning and assessment design and concerns of misinformation or biases in the outputs of the technology.


This scholarship project explores how this disruption may lead to enhanced learning, and skills development for Open University students by harnessing the benefits and identifying the risks for student of AI. A closed question survey was completed by 30 FASS students studying at level 2. Questions explored student’s attitudes towards AI as well as confidence in writing skills and isolation in study as issues that AI could address. Four of these students then participated in a workshop with a tutor in which they would devise an answer to a level 3 TMA question using AI during which the students’ qualitative feedback on the experience of using AI was collated, and later examined, by the researchers. Following this workshop, they completed an exit survey (with identical questions as the first survey). Following this research, we make three key recommendations including how communications to students regarding AI might be framed and how AI might be integrated into learning.


Key findings:

1. Students were hesitant about using AI as part of their studies and, initially at least, distrustful of the outputs of the technology. This included concerns over the accuracy of the outputs presented and the blurriness of plagiarism if using AI.

2. Students expressed concerns regarding their identity being repressed in using AI, recognising that the technology did not write in their voice or in words they would not use. Furthermore, because of their distrust of the technology, students would need to evaluate and verify the outputs before submitting any AI-aided work.

3. Following the workshop sessions, these students reported slightly more confidence in general writing skills in the exit survey when compared to the initial survey. Additionally, there were some slightly more neutral and even negative attitudinal responses towards AI when compared to the initial survey.

Recommendations

1. Promoting learner identity could be an empowering message for many students as assignments are an opportunity for their academic voice to be heard. Guidance to students regarding AI should emphasise the potential undermining of the learner’s sense of identity and agency from inappropriate (over)use of AI. As learners, one of the only visible actions and agency of a student is to complete and submit their own work as assignments. Emphasise that whilst challenging, preparing assignments can be interesting, satisfying and even enjoyable work getting to know and understand new ideas and information.

2. Communications should provide greater support for structuring a series of prompts to obtain a suitable output and emphasise the additional workload involved in using AI. It isn’t just copy-pasting an essay question and hitting enter but requires additional workload of verification, editing, re-wording and checking for relevancy. Student guidance (Notes) for assignments might contain templates for prompt engineering that also include reminders to critically assess the integrity and biases of the output.

3. Provide specific and contained channels for the use of Generative AI as part of learning design. Generative AI can then support types of learning – such as dialogical, ‘study buddy’ or developing an overview of the literature relevant to learning and assessment – which are more difficult for the OU to accommodate.

History

Collaborated with

  • Faculty of Arts and Social Sciences (FASS)

Sensitivity

  • Internal use only

Authorship group

  • Academic - Central
  • Academic - Regional/National (Staff Tutors and Student Experience Managers)

Institutional priority category

  • Achieving Study Goals
  • Students Learning Experiences
  • Other

Themes

  • Assessment
  • Accessibility

Subject discipline

  • Social Sciences and Global Studies

Usage metrics

    Faculty of Arts and Social Sciences (FASS)

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC