This blog post was co-written by Simon Buckingham Shum and Chris Girdler
UTS is a partner in an exciting multi-university research project listening to the Student Voices on AI in HE. The UTS/UQ/Monash/Deakin team has shared the first insights into the experiences of current higher education students as they navigate the intersections of Generative AI and their tertiary education journeys. This is the largest dataset on student views in Australia, contributing new evidence to inform wide-ranging practices and policies in teaching, curriculum development, feedback, and authenticity in assessment.
Highlights from the survey were just presented at the Microsoft ANZ HE Summit and the HEDx Future Solutions for Higher Ed conference by the lead team: Kelly Matthews (UQ), Margaret Bearman (Deakin), Michael Henderson and Tim Fawns (Monash) and Simon Buckingham Shum (UTS).
This project is unique, because the team has:
- spoken in-depth with students about their everyday uses and perceptions
- surveyed a diverse group of students from different subjects, disciplines and levels of study
- adopted a relational view of AI, in which students’ beliefs, understandings, backgrounds and learning conditions form a vital part of how we understand AI in use
The survey: exploring use, study and wellbeing
There have been many assumptions and claims being made about how students are using GenAI. To empirically investigate student use and beliefs of GenAI, the project team conducted 20 focus groups with 79 students from four Australian universities. They then created a comprehensive survey to capture large scale data (>8000 students) specific to the Australian higher education context.
The survey’s approach was to understand how and why students are using AI with 3 focus areas:
- Use – perceptions, use, access, usefulness, motivation
- Study – assessment, integrity, feedback, learning, future-readiness
- Wellbeing – emotions, trust, beliefs
Below are 3 takeaways from the survey results, with much more to follow. These comments and concerns echo what was expressed by students at our midyear ‘Student Partnership in AI’ workshops.
1. Feelings about GenAI are mixed
GenAI is a double-edged sword, and students are grappling with both sides in a way that can produce positive or negative emotions. The uncertainty and complexity that GenAI brings were evident in a wide range of responses to the question ‘How do students feel about AI?’ but these were balanced out my more positive, inspired emotions. The answers showed a fairly even spread of students feeling sceptical, worried, optimistic, grateful and excited – a far cry from being a blanket response.
2. High usage but low trust
83% of students responded that they were using AI for their studies, and 44% on a weekly or daily basis. The equity-building potential of GenAI tools was shown by 18% of respondents using it to help with accessibility challenges, but much more can be done here. The high level of AI usage may be eye-opening to some – but around half the students don’t trust what AI generates, with only around a quarter trusting what AI generates. Despite this lack of trust, only a quarter of respondents reported that they never or only sometimes change the original output of AI.
3. More guidance and transparency are needed
With so much uncertainty, students are seeking guidance from universities to navigate the changes instigated by the greater presence and usage of GenAI. The vast majority of students (91%) are concerned about breaking the rules, while 40% have indicated that they’ve used AI in assessments when they were not supposed to. A lack of guidance was highlighted as a particular pain point, with low numbers of students responding that their institution gives them enough guidance to use AI effectively in their studies (32%) and for their future profession (23%).
So, that’s the overall story — we’re now crunching the numbers for UTS students.
Do any of these insights surprise you? Or do they echo what you are experiencing in the classroom?