Interview with John E. Dolbow, PHD, Assistant Vice President for Research at Duke University’s Office for Research and Innovation.
- You have been appointed as Assistant Vice President for Research to improve data management practices at Duke and I’d be curious to learn what would you like to see in five years from now, in terms of data management?
John Dolbow: As the Assistant Vice President for Research within the Duke Office of Research and Innovation, I lead the Research Data Initiative. The Initiative has three aims: 1. to facilitate efficient and quality research; 2. to ensure data integrity and 3. to foster a culture of data sharing. In five years, I hope that the Duke culture around data sharing will change significantly, so that sharing research data by Duke Faculty and research scientists will be the norm rather than the exception. And, I hope that we would be in a place where the life cycle of research data is well understood on campus and there's a general understanding of the viable options for data storage and retention as well as support for data curation. Finally I'd hope we are in a place where researchers feel that the university is providing centralized resources that genuinely help their projects thrive.
The Research Data Initiative is now working to review the existing resources and the practices and policies around data security, data management, data retention and analysis across the entire institution. We're trying to identify opportunities for Duke to increase its ability to efficiently and effectively manage existing risk in this space and we're evaluating and presenting recommendations for meeting funding and publisher requirements for research data. We're beginning to scope out the procedures and resources that will be needed to support a new policy. Duke’s policy on research data was last revised in 2007. A lot has changed since then. Given the current research landscape and the funding requirements, we're actively engaging the community to figure out what Duke’s policy on research data should be.
The draft policy is really the first key milestone for the initiative and we hope to disseminate it to the Community for public comment in Fall 2021. We will then revise the draft based on community feedback and finalize the policy. This is the second key milestone, and that's when actually the real work of the Initiative will begin in earnest, because we'll be looking to set up procedures and resources to support research faculty and staff on Campus and the School of Medicine, in accordance with the new policy.
”It is really vital that we obtain buy in from our community and support for this initiative”
- How will the Duke community be directly impacted by this policy? Will researchers be required to transition to electronic research notebooks or to share their research data?
JD: There will be some requirements that come out of the policy, but it can't be a one size fits all type of policy. We will have different areas around retention, storage or access. We are looking to give researchers different options that they can select in order to meet requirements. The new policy will apply to all Duke, but it needs to be flexible enough to provide options regarding how to be in compliance with the policy.
- Do you foresee any implementation challenges or resistance to change and have you fought how to overcome them? Engaging the research community to comment and provide feedback to the draft policy is a very important step in this respect, by bringing people on board to embrace and contribute to this change.
JD: Yes, it is really vital that we obtain buy in from our community and support for this initiative. We're actively engaged and socializing this initiative: I am giving presentations in front of departments, having a lot of conversations with research leaders around campus and the School of Medicine to orient them to the fact that we are developing a new policy and also trying to understand what kinds of procedures and resources will be needed to support the policy. Research data is like an octopus with tentacles that reach into all areas of the university life, ranging from pure procurement, such as purchasing a particular measurement device, all the way to research publications that come out of research studies. A real implementation challenge is understanding what needs to be done in order to retain data into the future for some period of time. We are trying to effect a cultural change around research data, and we anticipate there will be some resistance. Our task is to convince the Duke community that there is real value in ensuring the integrity of their research data retention practices, as well as sharing their research data within their community. Our goal is to structure facilities and resources that make things easier for researchers and facilitate high quality impactful research rather than burdening them.
As the initiative moves forward we'll look to highlight examples of faculty from around the university who've done this successfully. We want to engage the trainees who are an important part of this initiative. Students and research fellows in many cases bear the burden of needing to follow new guidelines and protocols around research data, whether they come from the sponsors, or they come from the institution. But they're also, in many instances, the drivers of real innovation and change within research labs. Providing excellent training to them around research data is vital for the Research Data Initiative and for the goal of changing the culture on campus.
“Science occupies the vast majority of my time”
- How did you get where you are today in your career?
JD: I've been at Duke for more than 20 years now. This is actually my first academic position and I started here almost immediately after defending my doctoral dissertation. I think I defended in late July and started my position in the Engineering School here at Duke on August 1st. As I rose through the Faculty ranks at Duke I was very fortunate to have been given some leadership opportunities along the way, ranging from being the Chair of the Engineering Faculty Council, for example, to the Director of Graduate Studies in Civil and Environmental Engineering, and later in Mechanical Engineering and Material Science. A few years ago I joined the Rhodes Information Initiative, a highly interdisciplinary initiative at Duke where I had the opportunity to interact with colleagues interested in data science, from engineering faculty to computer scientists and mathematicians. I got to know Larry Carin our former Vice President for Research who encouraged me to consider the Assistant VP for Research position. I jumped at this opportunity because I understood that in this role, my work would have a broad impact across the university.
- Could you tell us about the focus of your research and its social impact?
JD: Science occupies the vast majority of my time. I still have a fairly large and active research lab. I'm a computational scientist who studies fracture mechanics. My lab attempts to develop computer models of systems that are capable of predicting whether or not systems will fracture in response to applied loads. My lab is involved in several projects that have real social impact, the first I'd mentioned is a project being led by Lawrence Livermore National Laboratory concerning carbon capture and storage. The project is focused on sequestering CO2 in abandoned wells and forecasting their structural integrity far into the future. In my lab we're helping to develop computer models that can predict how CO2 stored in the subsurface at high pressures will react with wellbore cement over thousands of years and the propensity for the cement to fail due to the formation of micro scale fractures and their possible growth.
The second project that I'm involved in that has a societal impact concerns Laser Lithotripsy. This is a medical procedure to break up large kidney stones in patients and it works by ablating them with a laser that's mounted to the end of an endoscopic probe. My lab is part of a center on Laser Lithotripsy at Duke; we help to develop improved treatment protocols by simulating some of the fundamental physics involved.
- How have mentors helped you succeed?
JD: I was very fortunate to have had excellent mentors over the course of my Duke career.
In particular as an Assistant Professor starting out in this world, in academia, my career was influenced in fairly significant ways by Tod Laursen who chaired the committee that hired me at Duke. Tod is also a member of my scientific community. He is currently serving as the acting President for SUNY Polytechnic Institute in New York. He was simply an enormous help to me at the start of my career really showing me the way from helping to identify strong PhD students in the pool of applicants to securing funding for my research. Tod was there if I needed his help on just about anything, and I really can't thank him enough. Tod remains a dear friend and colleague. These days I continue to have excellent mentors around the university. Geeta Swamy is really serving as somebody that I think leads by example, someone I admire a great deal and look to emulate, to the extent that I can.
“It is not a choice to be productive versus sharing the science. There's a way to share your science which may actually lead to more connections, more productivity and ensure better integrity of your research.”
- What are some of the challenges to research integrity in your area of research?
JD: Many of my colleagues are fundamentally focused on the development of new algorithms and in order to demonstrate the potential for those algorithms they often rely on research codes that they developed internally within their own labs and their own groups, as opposed to relying on open source research or research codes that are shared in any way. This can present a real challenge to researchers who are looking to reproduce their results and build on them because they may not necessarily have access to the same set of research codes. I think it's a little bit embarrassing, to be frank, that we, as a community that's very computational, haven't figured out a way to better share our codes or to have contributed to a common codes database.
I think we're starting to see some changes. My own research group is looking to build and contribute to open source research codes. There are a lot of pressures on faculty just to produce results, and I think one of the cultural changes we have to try to effect in this space is to make people appreciate that it is not a choice to be productive versus sharing the science. There's a way to share your science which may actually lead to more connections, more productivity and ensure better integrity of your research.
- Why do you think people are reluctant to share their science?
JD: I think the burden in many cases is simply a lack of understanding: how to get started or what are some of the steps that need to be taken?. Another reason is that many PIs are willing to adopt approaches that are common amongst their wider scientific community, independent of Duke. Those cultural forces are stronger than anything that Duke might put in place. So, one of the things that we're looking to do is to identify people at Duke that are serving in leadership roles in these various scientific communities in order to change the culture in their broader scientific communities.
- How do you think we can best train our students to embrace the culture of research integrity earlier in their careers?
JD: Our orientation should be to identify good examples of researchers, students, and professors who have been successful with sharing their research results with the public and their scientific community, and to put a spotlight on those individuals, to celebrate their successes. If students can see the advantages of open science and of sharing research data, if they can understand how to do this very early in their careers I think they will be more likely to embrace these practices.
- What is your personal experience with open science?
JD: A huge change in my lab came about 10 years ago. We started to rely on an open source research code from Idaho National Laboratory, which is focused on the development of new nuclear energy technologies. The real advantages that we saw within my research group were that we could reference the code that we used to generate our results, which made it easier for colleagues to reproduce them. We're also patched into an entire network of people that were actively contributing to that research code and using that code right away. We made connections with colleagues working in different areas that we wouldn't have anticipated speaking to before. I think that's been a big success for us. We have a lot of confidence in our results. We've also been able to revisit some of our older studies. Sometimes you chase a new idea for a while and you might put that work to the side for a couple of years and you find yourself coming back to it, for one reason or another. But it can be a challenge to restart that work if your codes have essentially aged out of use. The nice part about working with an open source code is that this is something that's being actively maintained by the community. It is constantly being sustained and improved.
The other big advantage is that when an innovation is placed into that research code by a colleague of mine, we pretty much have access to it right away. And, vice versa, my colleagues can take advantage of some of the innovations that we've put in place in that code.
This interview is part of a series of interviews with renowned scholars, scientists and influential leaders invited to share their outstanding work and views on topics related to research integrity.
Read all the articles in the Research Interviews Series: