Archives

The “REPAIR” project is an initiative aimed at addressing systemic inequities in the way academic research funding is awarded, with a particular focus on improving fairness for marginalised early career researchers (ECRs). By partnering with key academic and research communities, including the Faculty of Brain Sciences and the Faculty of Arts & Humanities at University College London (UCL), REPAIR leverages diverse perspectives to reshape the (pre) selection and nomination processes.

At the core of REPAIR is a commitment to mapping out and understanding the current funding system to identify where biases might occur. This involves using diagrams to visualize decision flows and analysing academic networks to spot patterns of unfairness, ensuring that the areas most in need of change are targeted effectively. ECRs and other underrepresented groups are central to this process. Through interviews and direct engagement, REPAIR will gather personal experiences and insights, making certain that the proposed changes address real-world challenges and lead to genuine improvements.

Collaborative efforts with partners like Voices of Colour, ALBA, the British Neuroscience Association ECR networks, and the UCL Research Culture Team will help develop new strategies to reduce bias. This includes creating workshops and innovative tools to make selection criteria fairer and using software to monitor the diversity and inclusiveness of the applicant pool. REPAIR will rigorously test these strategies in settings like the annual preselection process for UKRI fellowships, using both quantitative data and qualitative feedback to measure their effectiveness.

How can internal (pre)selection processes for research funding and key career decisions (e.g., prize nominations, industry knowledge exchange) be redesigned to minimize bias and better support marginalized early career researchers?

Objectives

1. Existing System Mapping: Review prior studies and collaborate with stakeholders to identify disparities in information, resource access, and support, and map potential biases in social promotion (using social network analysis for internal research prizes and grants).
2. Amplifying Silent Voices: Use qualitative methods like focus groups and interviews with
   marginalized ECRs and other key stakeholders, including those who have left academia.
3. Co-design and Implementation: Utilize findings from (1) and (2), along with existing
   literature and partnerships with research, culture, and EDI networks, to co-design,
   implement, and evaluate interventions to reduce these biases and disparities

Funding

The project is funded by ESRC via the https://edicaucus.ac.uk/.

People

This project team consists of:

• Dr Christin Henein (PI), Institute of Ophthalmology, UCL
• Prof Aikaterina Fotopoulou, Clinical, Educational & Health Psychology, UCL
• Prof Anna Cox, UCL Interaction Centre, UCL
• Dr Simona Aimar, Dept of Philosophy, UCL
• Dr Naaheed Mukadam, Division of Psychiatry, UCL
• Dr Natalie Marchant, Division of Psychiatry, UCL

With partners:

• ALBA Network;
• the British Neuroscience Association;
• The European Society for Cognitive and Affective Neuroscience;
• the British Academy’s Early Career Researcher Network;
• Voices of Colour;
• UCL’s Institute of Advanced Studies and Early Career Network in Arts, Humanities and Social Sciences;
• UCL’s Research Culture team;
• Centre for Equity Research, Faculty of Brain Sciences, UCL;
• Research Coordination Office for Life and Medical Sciences, UCL; and
• the UCL Neuroscience Careers Network (NCN).

Publications

Are you an academic who is using LLMs in your work? We’d love to hear from you. Complete this survey and we’ll send you a token of our appreciation or find out more about this project here.

Academic jobs are demanding, intellectually intense and stimulating, and attract high performing individuals. They incorporate research, teaching, institutional and disciplinary service, and knowledge exchange. However, despite this well known description of the job, there are many aspects of these roles that are not well understood, under-recognised and unequally distributed. This project delves into the challenges faced by academics, particularly women and non-binary individuals, who bear the burden of invisible labour in the academy due to increased emotional labour, as well as teaching, service, and administrative tasks, when compared to their male colleagues. These non-research activities often go unnoticed but significantly impact academics’ time for research-related work and career progression.

We propose using Large Language Models as a potential solution to help alleviate some of the invisible labour, allowing them more time for research and non-work activities.

Funded by UCL Research Culture, we are currently undertaking research to explore how LLMs can support academics in managing their invisible labour, ultimately creating a more equitable academic environment. Take part in our survey.

Research Questions

1. What is the nature of the invisible labour carried out by academics?
2. What are the current practices of academics using LLMs to support their invisible labour?
3. How might future tools based on LLMs be designed to support academics with their invisible labour?

What other research are we doing in this space?

We are:

• developing ResearchComplianceBot that helps academics with the research compliance process ie applying for data protection registration and ethics clearance.
• exploring how AI can help time-poor academics to conduct peer review

People

This project is being developed by Jon Mella, Sarah Frampton, Prof Anna Cox Dr Anna Dowthwaite, with assistance from MSc students Yvonne Chang and Anya Emmons.

Publications

Mella, J., Frampton, S., & Cox, A.L. (2024) Unseen Work: Leveraging Generative AI for Invisible Academic Labour. MCI-WS09: Workshop on Generative Artificial Intelligence in Interactive Systems: Experiences from the Community

Cox, A. L., & Mole, S. E. (2024). Five questions on improving diversity, equity and inclusion in UK bioscience research or “How can UK bioscience be changed so that those from marginalised groups can thrive?”. BBA advances, 5, 100114.

This project aims to make dementia research more inclusive that addresses the challenges experienced by researchers and ethnically-minoritized communities using co-production methods.

We urgently need greater participation of ethnically-minoritized groups in dementia research to understand the reasons for differences in vulnerability and survival, and to provide better interventions. We need to know what barriers ethnically-minoritized communities experience that prevent their participation in dementia research projects. We need to know what barriers researchers experience that prevent them from using inclusive recruitment practices. We need to know how to work collaboratively with multiple stakeholders to improve inclusive recruitment practices.

This project consists of three activities – focus groups, a survey, and a workshop. We will:

• Conduct focus groups with Black and South Asian community members to identify barriers to participating in dementia research
• Survey researchers about their current recruitment practices to identify barriers to inclusive recruitment
• Conduct a workshop, co-designed with community collaborators, that gathers researchers and other stakeholders together, to ideate and co-produce inclusive recruitment practices for dementia research

People and Funding

This project is being led by Dr Natalie Marchant and Prof Anna Cox in collaboration with Lynis Lewis NOCLOR, Ione Fraser, Lorraine Cezair-Phillip, Dr Naaheed Mukadam, Jabeer Butt OBE Race Equality Foundation, Dr Harpreet Sihre and Prof Katerina Fotopoulou.

Researchers contributing to the project include Shiping Chen and Elahi Hossain.

We are funded by the UCL Grand Challenge Justice and Equality.

The modern workplace often places demands on workers to quickly learn and apply new practices in their workflow to maintain productivity. One of the easiest ways to learn about best practices and solutions to common problems at work is via colleagues, particularly when using complex software such as spreadsheets, where expertise is essential to reduce the likelihood of errors. Motivating workers to make their expertise visible and to engage in knowledge sharing is crucial for such interactions to occur, however, current research offers limited perspectives on the experience and motivations of knowledge providers.

Focusing on the context of the spreadsheet, this project aims to develop our understanding of motivational barriers to knowledge sharing behaviours among workers, and then to develop and test interventions which could support workers to engage in more effective knowledge sharing behaviours to improve learning within their community.

Funding

The project is funded by an iCase studentship.

People

This project is being delivered by Nancy Xia under the supervision of Prof Anna Cox, Prof Duncan Brumby and Dr Advait Sarkar, Microsoft Research.

Publications

Exploring Accessibility and digital system support for neurodivergent students

Blended learning has become a mainstream learning experience for many universities, including UCL, during the COVID-19 pandemic. While blended learning has received much attention over the years, particularly in the area of education, we still know little about how people with disabilities engage in blended learning from their homes and what access means in this context.

To understand and rethink accessibility in blended learning, we proposed a programme of study of blended learning practices of neurodivergent students who have Autism Spectrum Disorder, Attention Deficit Hyperactivity Disorder, learning disabilities (e.g., dyslexia) and psychosocial disabilities (e.g., anxiety, depression).  To understand the impact of university digital systems on neurodivergent students, we conducted an interview study. Led by a neurodivergent researcher, this study involved semi-structured interviews with 11 neurodivergent students at UCL. These students, diagnosed with ADHD, autism, or both, shared their experiences and challenges in navigating university digital platforms.

The interviews were designed to be open-ended, allowing students to discuss not just their difficulties, but also their coping strategies and suggestions for improvement.

We used a thematic analysis approach to dive deep into their narratives, identifying common themes and unique insights. By centering the voices of neurodivergent students, our goal was to uncover the real-world impacts of these systems and to highlight practical changes that could make a significant difference in their academic lives. The analysis revealed three key themes:

• Neurodivergence in academia is not experienced as advantageous, but as an invisible disability}: Participants reported various symptoms of their neurodivergence that impacted academic work, relationships, self-esteem, and more. While there is a common perception of neurodivergent “superpowers,” neurodivergent traits were described as leading to negative feelings, mental health challenges, and worse symptoms.
• Complex digital workflows trigger symptoms, provide negative user experiences, and impact learning}: Excess user workload emerged as a major concern, encompassing both physical and mental workload. Participants highlighted challenges related to limited integration of systems, number of steps in sequences of interaction, and interaction issues related to lack of simplicity, low intuitiveness, low consistency, and automation. Visual features, customization, and video-related elements also influenced user experiences. The digital systems were reported as triggering symptoms, such as procrastination and emotional dysregulation, negatively affecting academic progress. Login and password systems were a particular source of frustration, demanding urgent attention for improvement.
• Students implement coping strategies to help them manage their personal and academic experiences: Participants shared their systems for managing their experiences better, including digital tools, physical items, creating a sense of urgency, and accountability buddies. Acknowledging and integrating these strategies into system design has potential for fostering a more supportive environment.

Funding

This project was funded by the UCL Centre for Equality Research in Brain Sciences

People

This project was led by Prof Anna Cox and Dr Anna Dowthwaite with assistance from Alex Tcherdakoff and in collaboration with Dr Paul Marshall and Dr Jon Bird.

In recent years, new data protection laws have put more responsibilities on researchers covering how we gather, store and process participant data. Many researchers find these requirements confusing and it’s easy to become overwhelmed with the sheer amount of legalese one has to read to truly understand data protection laws and its practicalities.

At the same time, researchers have to obtain ethics approval from their institution and make sure that their project complies with any associated requirements.

This is why we are designing a tool called Research Compliance Buddy. This tool will guide researchers as they fill in applications to register their research with the Data Protection Team and the relevant Ethics Committee. It will help them decide which aspects of applications are relevant to them and which are not, to quickly understand jargon, and guide them directly to the information they need to read to best plan their particular type of research.

We have conducted semi-structured interviews with a diverse group of researchers within the Division of Psychology and Language Sciences at UCL. The data collected during these interviews has been analysed and is guiding the first stage of the design of a chatbot.

People

This project is being developed by Dr Anna Rudnicka, Prof Anna Cox and Elahi Hossain with assistance from Elizabeth Sahadevan.

Publications

Hossain, E., Dowthwaite, A., & Cox, A.L. (2024) ComplianceBot: Leveraging Generative Conversational AI to Support Researchers with Academic Compliance. MCI-WS09: Workshop on Generative Artificial Intelligence in Interactive Systems: Experiences from the Community

In this project we focus on alleviating the peer review crisis in academia. With an increasing number of submissions and a limited number of reviewers, reviewers are becoming overburdened, leading to delays in the review process due to a shortage of reviewers. This research explores using AI tools, such as ChatGPT, to support reviewers by reducing their workload and improving efficiency without compromising the quality of reviews.

The project involves qualitative studies to understand the needs and demands of both reviewers and authors. The initial study examined how reviewers conduct peer reviews, identifying key challenges including iterative interactions with authors, learning how to review, review complexity, lack of training, and workload issues. Another study explored authors’ requirements and attitudes, emphasizing the need for good feedback, optimism about AI, concerns about trust, ethical considerations, and the impact of AI on review processes.

The research also includes an experiment using AI (ChatGPT) to reduce reviewers’ workload and enhance review quality. Reviewers perceived both benefits and challenges, such as saving time, providing different perspectives, and improving confidence, while also facing issues like initial setup time and integration with current workflows. The project underscores the necessity of effective collaboration workflows and robust data protection measures for implementing AI tools in the review process.

People

This project is being developed by Shiping Chen, under the supervision of Prof Anna Cox and Prof Duncan Brumby.

Publications

S Chen, DP Brumby, AL Cox (2024) How to Alleviate the Peer Review Crisis: Insights from an interview study CHIWORK2024

S Chen, DP Brumby, AL Cox (2023) How to Alleviate the Peer Review Crisis: Insights from an interview study CHI 2023 workshop “In2Writing: Intelligent and Interactive Writing Assistants”