AI for all - BAG meeting - 10th September 2025 by Maria Canal

AI for all - Technologies for Inclusive Bioscience Education

mqbssmc3 — 2025-08-26 10:05:24 UTC

A very warm welcome to the Bioscience Awarding Gap (BAG) Network at AI For All. In this meeting we will explore opportunities, share ideas and discover practical ways to adapt our bioscience teaching to make the most of these powerful new tools. The meeting aims to equip educators with the knowledge and tools to turn ideas into impactful action.

Meeting Hubs

mqbssmc3 — 2025-08-26 10:08:01 UTC

A total of 8 local Hubs plus a virtual one will be hosting this multi-hub hybrid meeting:

Kingston University
Liverpool John Moores University
Middlesex University Hub
Royal Holloway
Sheffield Hallam University
University of Leicester
University of Manchester
University of Westminster
Virtual Hub - for those attending on their own

Who are we?

mqbssmc3 — 2025-08-26 10:17:00 UTC

This meeting is organised by the Biosciences Awarding Gap Network, an advisory group to the Heads of University Biosciences (HUBS). The group promotes within HUBS the communication, discussion and adoption of practices that are shown to reduce demographic awarding gaps in the biosciences.

Members of the BAG Network Steering Committee:

Dr. Amara Anyogu, London Metropolitan University
Dr. Maria Canal, University of Manchester
Dr. Nick Freestone, Kingston University
Dr. James McEvoy, Royal Holloway University of London
Dr. Prachi Stafford, Sheffield Hallam University
Prof. Ian Turner, University of Derby

Feedback please!

mqbssmc3 — 2025-08-26 10:50:58 UTC

Please let us know what you thought of the meeting by completing the Menti above or by adding your comments below. Your feedback is much appreciated!

2025-09-10 09:52:35 UTC

1. Incorporate AI spelling, grammar and clarity check as part of written assessment brief.

2. Institutional adjustments – support plans, physical adjustments (adjustable benches) etc.

3. Differentiation between genuine disabilities and stretched capabilities.

Reasonable adjustments.

4. Bespoke help – alternative assessments etc.

percyl4 — 2025-09-10 09:52:38 UTC

Appropriate background experiences – education, work etc for courses/program
Carefully designed programs/modules
Evaluations of student knowledge
Learning Resources, inclusive
Understanding student cohort
Listen Studentvoice
Staff approachable – ‘safe space’
Student communities – for support for personal, mental support, reduce isolation/loneliness
Appropriate language/explanation/for familiarisation
Expectations for assignments
Study skills to support are offered
Asking students what they want/suggestions (mid module for support) – adjust then consider impact
‘Role models’ –students/staff that students connect with/alumni connections
Material/content of modules is relevant/appropriate examples/ videos/(how learn will be different)
Assessment choices – varied, flexible/alternative assessments
If field trips/visits/laboratory practicals – all need to be accessible – all equipment needs to be accessible
Consider EDI /reasonable adjustments/not all students will declare
Student guidance for being supportive of others – group work
Student training - good use of representatives
Appropriate staff guidance/reflection

Accessibility to AI tools

mqbssmc3 — 2025-09-10 09:52:41 UTC

2025-09-10 09:52:58 UTC

2025-09-10 09:53:09 UTC

2025-09-10 09:53:21 UTC

Room 14

2025-09-10 09:53:29 UTC

Team based learning. Balanced teams based on gender, subject taken on A level.
Formative assessments to address knowledge variation of students.
Pre-reading material can be used to address prior knowledge variation.
Academic coaches/strategic coaches can help to address knowledge variations too.

Do assessments suit diverse learning strength? Yes, they do and support is provided. Assessment are in form of MCQ or OSCE. University owned ipad is used which has a virtual learning space to use the assessment.

Are all students are equally supported to succeed? Yes. By providing support in forms of 1to1 meetings, and success/academic coaches. Student disabilities are all considered and reasonable adjustment around it is provided.

Who might this module unintentionally exclude. Modules are designed in an inclusive way. Unless a student has a disability or a reason which they may not have declared, all students should feel supported.

2025-09-10 09:53:47 UTC

Room 12 (we think) CCCU + Edge Hill)

2025-09-10 09:53:59 UTC

Prior Knowledge variation: start very “basic”/general at L4. Quizzes at the beginning of the session to gauge prior knowledge.

Do assessment suit diverse learning strengths? Variety of assessments (essays, write ups, posters, exams, online assessments, critical reviews, leaflets, etc. Sometimes constricted by university regulations.

Who may unintentionally be excluded? Physical disabilities can be a challenge with practical degrees. Fieldwork. Physical adjustments are often in place, but there are challenges with staff arrangements. Literacy in AI use can be different between staff/student. Adjustments may not be always obvious. Learning Plans do not always reflect what happens in science sessions (labs specifically); some support plans can be extremely generic (staffing is very low).

Are all students supported to succeed? Staff shortages, pinpoint what support students need, disclosure issues. Students previous experiences of school not always known.

RHUL sheet

2025-09-10 09:54:15 UTC

2025-09-10 09:54:17 UTC

Room 10

2025-09-10 09:54:30 UTC

We were concerned about how we can best help our students to understand the limitations of AI in their course work.

Also, AI is here and we have to work around that.

2025-09-10 09:54:40 UTC

2025-09-10 09:54:53 UTC

From Middlesex University - one idea we had was to integrate timetabled time in the module for students to carry out assignments - this would ensure students who cannot work from home or need extra support can complete the work and feel supported

2025-09-10 09:57:00 UTC

Breakout room discussion

2025-09-10 09:57:32 UTC

Module/Institutional level

Middlesex Uni - Setting ground rules, expectations, background for students from different organisations.
Assessments - Variation from traditional means of assessment styles.
Presentations rather than written reports and graphical abstracts to be more anti-AI.
By making sure that we give them any support they require in their learning.
We could test their learning by formative assessments
We could check their progress and how they are performing in formative and summative assessments.
By monitoring their engagement in different teaching activities

Assessment styles

Essays, Lab reports, Group presentations
Infographic
Poster presentations
Case-studies problem based questions
Systematic Reviews
UG Research Thesis
Critical Reviews
There are different types of assessments to cover different strengths of the students
Some students prefer Group work and some students they do not.
Some students prefer Exams some others course work
Laboratory simulations

Lobster simulations or Learn Science Simulations

Marketing recruitment challenges due to weaker candidates being more proficient at getting to interview stages.

2025-09-10 09:58:11 UTC

How is prior knowledge variation addressed? We should challenge assumptions about what students from diverse educational backgrounds can achieve. In some contexts, differences are seen, but this is not always the case. We should address prior knowledge variation head-on by discussing this with students and giving them the space to share their anxieties and needs. We also discussed the challenge of how some students navigate support, such as study skills. The students who attend gain a lot from it, but those who need it most often do not attend for various reasons.

mqbssmc3 — 2025-09-10 09:58:16 UTC

How to get students to access the support available?
Group work is beneficial for awarding gaps, but how do we effectively manage it, both from staff and students points of view
The importance of assessment literacy
How to best support international students and those whose English is not their first language?

2025-09-10 09:59:20 UTC

2025-09-10 10:00:59 UTC

2025-09-10 10:01:43 UTC

2025-09-10 10:02:39 UTC

Leicester - Table 3

Using data

mqbssmc3 — 2025-09-10 10:08:33 UTC

How we can best use the data available (e.g. attendance, student engagement, etc.) to better understand if the correlation with student success, and to understand what is stopping our students and how we can best support them

2025-09-10 10:09:42 UTC

Middlesex University's thoughts

Working out what success means

2025-09-10 10:17:07 UTC

We felt the first step was to define success. It may mean different things to different groups of students. Is passing enough? Remaining on a masters/with IE programme?

Using data to find who we are failing

2025-09-10 10:19:22 UTC

We felt it would be important to use attendance and attainment data to identify the groups who are underachieving. These may be different at different institutions so relying on others’ research, while giving useful pointers, may not be appropriate in all cases.

Solutions

2025-09-10 10:22:42 UTC

“Solutions” also have the potential to exacerbate problems. Unfortunately, the interventions we might use tend to have a long lag time before we start to see results. It takes at least a year to implement a new form of assessment in our institution, and in the current environment in which AI is evolving so quickly, this is not quick enough.

Ideas

2025-09-10 10:44:32 UTC

To ensure all students succeed, we can build on practices already in place while introducing inclusive strategies. We discussed during Welcome week, students complete essential lab skills, work in groups, and analyse each other’s techniques, which helps establish a shared baseline of competence and fosters peer learning. At CCCU we use portfolio of competencies to build skills and confidence in the lab from the beginning. In addition, exploring alternative assessments, such as providing a journal paper prior to the exam and asking students to critically analyse it during the exam, encourages critical thinking rather than memorisation and accommodates different strengths. By combining these approaches with ongoing feedback and multiple assessment pathways, we create a supportive and equitable environment where every student has the opportunity to thrive. Soft skills are something employers are looking for in graduates and so this can be challenging for students with a mixed ability/strength.

2025-09-10 11:13:51 UTC

Middlesex University HUB

UoM SBS

mqbssmc3 — 2025-09-10 11:17:34 UTC

Training for students and staff on how to use AI tools to leverage benefits. Ensure that all students have equal access to AI tools through institutional licence.

Leveraging the power of AI

2025-09-10 11:17:54 UTC

We suggest using AI to analyse data to identify why the attainment gaps exist. Uploading data on student performance, characteristics, unit assessment methods, delivery methods etc, could allow patterns to be identified beyond those that human analysis might identify.

2025-09-10 11:21:57 UTC

percyl4 — 2025-09-10 11:22:15 UTC

Use AI to help us understand what is causing the awarding gaps? (For example, deep analysis of student surveys, progression and award data) Better ways to use the data?
Support students in how to use genAI - be critical – trust of AI?
Disparities in use of AI, some consider it ‘cheating’ others rely on AI – adapt AI to align these?
If financial issues – could Universities pay for the most up-to-date AI for all?
AI - use for language – can we correct these terms?
Grammarly or appropriate apps can support with writing, especially for those not confident with this
GenAI creation of ‘podcasts’ of literature – to support explanation of literature articles
GenAI tools for conversion of text, eg assignment briefs, lecture notes slides to different languages
Appropriate genAI can be critique work (particularly for 1^st gen students with limited/reduced support).
Equitability between courses/Universities on use of AI
Support with data coding for all
Could be used for Q and A style

Room 8

2025-09-10 11:22:28 UTC

1. Levelling the playing field with grammar, spelling and clarity amendments.

2. AI generated podcast or alternative forms of delivery.

CCCU + Edge Hill

2025-09-10 11:22:54 UTC

Can unis pay for a paid subscription, to level the plain field? Do we have a set of “recommended” tools for the uni, so we can direct students to those tools (and get training ourselves).

Training in the use of AI, both for staff and students.

Training for staff, so we can support students better (or the gap will only widen). The training needs to be consistent, not only one offs

It can make your writing more academic/polished, but some worries about losing individuality

2025-09-10 11:23:08 UTC

2025-09-10 11:23:12 UTC

Didn't catch room sorry

2025-09-10 11:23:15 UTC

2025-09-10 11:23:36 UTC

2025-09-10 11:23:56 UTC

RHUL sheet

2025-09-10 11:24:12 UTC

2025-09-10 11:25:00 UTC

2025-09-10 11:25:52 UTC

Teaching students prompt engineering - how to develop and refine prompts (recognising confirmation bias so designing prompts in balanced ways for balanced answers) and developing resources or content that is more accessible for themselves

2025-09-10 11:26:04 UTC

Virtual Room

2025-09-10 11:30:58 UTC

There is an assumption that most academics are confident using GenAI, but this is not the case. Many students are more AI savvy than we are, so making space for conversations about how they use AI would be helpful. We considered whether AI is taught in school, and if this could help or exacerbate 'AI literacy' divides. AI has a great role in reducing awarding gaps, and the short answer to all the discussion prompt questions is yes. Still, we need to be intentional and collaborative (work with students) on the outcomes we want to achieve, which tools are appropriate, and how to use them. We also need to have honest conversations about how students are currently using AI. The assumption seems to be that they are doing something wrong, but we don't know.

AI toolkit impact evaluation

mqbssmc3 — 2026-05-03 10:26:45 UTC

Please share your name and institution, how you are using (or plan to use) the AI Equity Resource Toolkit, the module(s) or programme(s) involved, and—if applicable—any reflections on how it has worked in practice, to help us evaluate its reach and impact.