On Thursday 28th of April, Early Career Researchers (ECRs) from the mental health data science community came together to talk about building collaborative teams for Mental Health Data Science.
We are collaborating with HDR UK’s new hub - DATAMIND - which is working to enable researchers and stakeholders to find and use UK’s rich datasets and health records for mental health research.
The talks and discussions highlighted some of the barriers to collaboration in the Mental Health Data Science field as well as opportunities for collaboration and career development, particularly for ECRs.
Here are our top four takeaways from the event.
1- Funding for Collaboration: Persist, accept rejection, and be entrepreneurial
2- Define what a successful collaboration would look like for you and ask for it
3- Aim big, but start small- play the “long game”
4- Even the collaborative playing field
Top takeaway #1 – Funding for collaboration: Persist, accept rejection, and be entrepreneurial
Compared to other disciplines, Mental Health Research has lagged behind on both collaboration and funding fronts. In fact, during the discussions, ECRs have highlighted how limited funding streams hamper collaboration with other disciplines. Right now, there is a gap in the flexibility and diversity of funding hindering the integration of Mental Health Data Science Research. So, how to overcome the funding barriers?
While there is no magic recipe, our keynote speaker, Rob Stewart (King’s College London), advised ECRs to “persist and be entrepreneurial” in their funding applications. Researchers need to cast a wide net and keep an eye out for opportunities beyond mainstream funding sources.
As Ann John (Swansea University) phrased it, “you are going to get rejected… it is not about you, so try not to hold it to heart and keep going”. There are different and random factors that interplay in funding decisions. One way to explore these factors is to join funding panels. While the process may be tedious and time-consuming, it gives ECRs some insight into how decisions are made, the evaluation criteria, and how to write more robust proposals.
Top takeaway #2 – Define what a successful collaboration would look like for you - and ask for it
Our guest speakers, Deepti Gurdasani (Queen Mary University of London) and Giouliana Kadra-Scalzo (King’s College London), gave remarkable examples of innovative and collaborative projects in Mental Health Data Science and ways to approach collaboration in this field. The challenge is often how to begin; what is the first step?
This could be intimidating to many young scholars. So, let’s think beyond the buzzword. Giouliana Kadra-Scalzo’s recommendation is to start by establishing who you want to work with, the collaboration you are seeking, and how it will drive your research forward. This makes it easier for researchers to identify people and groups they want to work with within and outside their institution and reach out to them. More often than not, people would be open to a conversation.
While you may think that your interest in forming research partnerships is common knowledge, this might not be true. Talk to people in your network, and while there might not be opportunities for collaboration at that time, they can be catalysts for future collaborations. When the opportunity arises, “you will be on the top of their mind”.
Top takeaway #3 –Aim big, but start small- play the “long game”
The lack of integration throughout the research community hinders progress in the Mental Health Data Science field. Scientists and organisations need to work together better to understand mental health and ultimately better prevention and treatments.
However, this is not an easy task. The Mental Health Data Science research community is comparably small, and time is a significant limitation for researchers. Our Keynote speakers, Ann John (Swansea University) and Rob Stewart (King’s College London) agreed that you don’t need to start big- it can be people within your team. Go to networking events, workshops, meetups, and conferences and grab on to opportunities even when they do not fit your ideal role. Then, build on that. The responsibility is not on one individual but on the whole Mental Health Research community to engage with each other.
Collaboration is a process that needs nurturing and is inherently upheld by trust. Building this trust is essential; people you meet early on in your career may simply become partners and collaborators years later. There are no guarantees, however. Networking does not always lead to another project. What it can do, is “open a door in your mind” that might change the questions you’re asking and your approach to answering them. Simply, it is scientists talking to other scientists. But, if, like most of us, you do not feel comfortable approaching others, Ann’s advice is “play to the strength of people in your team. You do not have to do it all... no one expects you to go live on Panorama”. Small strides will get you there.
Top takeaway #4 – Even the collaborative playing field
In the absence of infrastructure to govern research partnerships, collaboration often occurs in a random and ad hoc manner rather than long term. There are many barriers to long-standing collaborations which hamper progress. Overcoming these barriers is essential to maximise the value of datasets for better mental health collectively.
Getting a highly skilled group with different backgrounds and disciplines to work together is challenging. ECRs shared how the process is not straightforward and can get messy. Since the research project is often unpredictable, they often find themselves over committing or under committing. Given the lack of structure, it is vital to agree on defined project roles, responsibilities, milestones, and partnership tools (meeting plan, communication platforms) from the onset of the project. Steer clear of jargon, and if you don’t understand, ask.
Another barrier is the unequal access level to data between collaborators. Since different collaborators will be responsible for generating data, analysing, and interpreting results, every step of the process should be documented and support findable, accessible, interoperable, and reusable (FAIR) data standards. DATAMIND can help with this. DATAMIND provides a central, integrated data infrastructure with FAIR mental health datasets. It has a federated approach offering a trusted environment that makes data visible. Although datasets are hosted in their own servers, DATAMIND will make it easier to know what information is out there and what datasets of cohorts have asked the questions you are interested in.
