Data scientists don’t often get splashy headlines, and so it’s possible even some DGHI veterans may not have heard much about the Research Design & Analysis Core (RDAC). But DGHI’s in-house team of methodologists is the quiet engine driving quantitative research across the institute, providing insights on research setup and data analysis that have fueled many successful grant applications, thesis defenses and high-profile research publications.
At the heart of the effort is Liz Turner, Ph.D., RDAC’s director and an associate professor of biostatistics, bioinformatics and global health who built RDAC from the ground up. Originally from the north-east of England, Turner worked as a collaborating biostatistician with the London School of Hygiene and Tropical Medicine (LSHTM) before coming to DGHI in March 2012. With fellow biostatistician Alyssa Platt, who had been hired a month earlier, she began offering expertise on quantitative methodology to DGHI faculty, creating the model for how RDAC operates today.
Now including 10 full- and part-time staff, RDAC has contributed to more than 150 grant submissions, resulting in $50 million in externally funded research. The team has also advised more than 300 students in DGHI’s Master of Science in Global Health program on thesis research projects and quantitative skills.
As part of our series of interviews marking DGHI’s 20th anniversary, we spoke with Turner about RDAC’s early beginnings and rapid growth – and why no statistician can save researchers from a badly designed study. The conversation has been edited for length and clarity.
You were recruited to DGHI specifically to start a biostatistics core. What appealed to you about that idea?
Well, the conversation really started to take shape around the summer of 2011. DGHI was still very young then, still really in its startup phase. And I think [DGHI founding director] Mike Merson saw having an embedded group of methodologists as a way to accelerate its research. I was at the London School for Hygiene and Tropical Medicine then, and Mike was on the board, so that became a natural connection for us to start talking. And it was exciting to me, because the mindset was very much, ‘Let’s try to build something and see where it goes.” Even when I shared that I didn’t really know much about the American grants system or the NIH, the answer was always, if we have the right people, we’ll figure it out.
It was also really appealing to me to be embedded in an interdisciplinary environment. LSHTM was wonderful, but it was definitely more of a traditional, public-health school approach. The walls between departments were pretty big, and I just felt like at DGHI I would use a lot of different skills. It was just the opportunity to be part of something new and to be a more well-rounded biostatistician.
Members of the Research Design &...
So when you got to DGHI, did you have to sell this concept? Or were there faculty waiting at your door?
There were, actually. I came in March 2012, and Alyssa Platt had started about a month earlier. John Bartlett, who was the associate director for research at the time, was just wonderful at helping us get started. He was so kind and encouraging, and he had a very good sense of how we could be working with faculty. We spent a lot of time meeting with people, but so much of it was just chance conversations, just bumping into people in the corridors. There was so much energy around the building in those days. You had the sense we were doing something new, and that our work was going to be impactful.
One of the first projects we worked on was Wendy Prudhomme O’Meara’s first NIH R01 grant, and she was a great advocate for our work. We also had some success with early-career researchers applying for K awards, and that became something we were kind of known for. So people did start knocking on our door. And then when Joe Egger joined [as RDAC’s associate director for education in 2015], we were able to expand and start working with the Master of Science students.
How do you approach working with students, especially those who haven’t had much exposure to statistics?
Yes, there can be some trepidation about working with us, especially when students are dealing with quantitative methods for the first time. It’s a lot of translation and interpretation. We try to get to know them and their research interests so we can use examples and narratives to make things more relatable for them. And we ask a lot of questions, because while we may have the expertise on quantitative methods, we don’t know all the areas of research that go on around DGHI. And that’s important for us to model for students, that it’s okay to ask questions when you need more clarity.
It’s such a critically important part of our mission to work with students. Over the past decade, we’ve had significant input on around a third of the master’s students’ thesis projects, and I think those students have really benefitted from working closely with an RDAC team member and seeing how these quantitative models work. But there are also the biostats courses, which have been taught by RDAC team members for the past 10 years, and so that’s another important way we are interacting with students
A lot of students come in dreading having to take that course. Do you feel like you’ve been able to win them over?
I think so. You know, the mythology about that course sort of gets passed down through the generations of students, and I think they come in expecting it to be harder than it is. But now, you have Chris Gray and Mercedes Bravo teaching the two core semesters, and Christine Markwalter running the biostats lab. And so there’s this trio of very person-focused women teaching those courses and bringing a lens of collaboration and transparency. And I think that’s the RDAC spirit, to make things as clear and simple as possible, and not to overstate or overcomplicate the work.
Even if they aren’t going to become quantitative researchers, we can help students be better consumers of quantitative information, and that means going beyond just looking at the data. We want them to have a sense of the skills and language they can use to interpret data and evaluate whether it actually supports what is being claimed. It’s about developing a healthy sense of skepticism about what they are reading.
What’s one thing you wish more people understood about research data?
That it’s about more than just analyzing statistics. I think people tend to think of us as statisticians who can bring a lot of specific skills in quantitative analysis, and that’s certainly true. But I always remind people that there are no fancy statistical tricks that can save you from bad data that is the result of a poorly designed study. We’re most useful when we can be in discussions from the beginning, helping PIs think about how to design questions and set up studies to get the data they need. And our PIs have been great about bringing us in early, often as they’re developing the research idea and well before they’ve even submitted a grant application.
And really, that’s the fun part for me – it’s understanding the human part of it and making connections with people so that we can work productively together. That’s why it’s so valuable that we are embedded within the community and that DGHI invests in giving us time to build those long-term relationships, because it’s so much easier to collaborate with people when you understand something about where they are coming from.
So really, good data is about human connection?
Sure, and maybe that’s part of what we’re trying to do, to defy the stereotypes about what working with quantitative methodologists is like. I think our PIs know we’re passionate about good research design and data analysis, and we’re not going to shy away from asking hard questions about their studies. But we’re always going to do it in a way that is open and human, and maybe even fun.
