Jump to main content

Faculty Research Spotlight

Dr. Chelsie Romulo

Image of Chelsie Romulo

Chelsie Romulo is a professor in Environmental and Sustainability Studies at the University of Northern Colorado. Her research spans several resource management contexts, but consistently seeks to understand what works and why to explain what contextual characteristics result in impacts and outcomes. She uses mixed methods approaches to integrate quantitative and qualitative data that can be applied to many different management and policy situations and frequently make use of existing data in new contexts. A former SESYNC graduate student pursuit member and Smithsonian-Mason Doctoral Fellow in Conservation, her dissertation research focused on community-based natural resource management in the Peruvian Amazon. Her current research interests delve into evaluating enabling conditions for payments for ecosystem services programs using big data machine learning models and developing assessment methods for teaching environmental and sustainability concepts. She is currently PI of an NSF IUSE grant using machine learning techniques as an assessment tool to understand how students learn complex sustainability topics in the Food-Energy-Water-Nexus.

Can you tell us a little about your current research?

I define my research as figuring out what works and why. My STEM Education research focuses on what works and why does it work regarding student learning. My NSF funded project is basically a proof of concept and first step for assessing learning. There exists “concept inventories” in many areas of science that test foundational knowledge, like force in physics, and we have concept inventories for climate change and geoscience.  These inventories are sets of questions that, up until this point, have been multiple choice type questions and usually each question is really focused on a basic concept.  So, what we're doing is trying to develop what we're calling the “next generation concept inventory”, where students can construct a response.  Commonly known as a short answer response, we can then look at how students can connect ideas.  We are working with a lab at Michigan State University called the Automated Assessment of Constructed Response Lab or AACR Lab, who has been doing this kind of thing for questions in sciences and engineering. They have not yet done this specifically for a concept inventory, and so we're kind of combining two existing processes - development of a concept inventory and using these machine learning tools to create an algorithm that will evaluate a constructed or short answer response.

The tool that we will end up with at the end of this could be applied in any classroom - we are starting with food, energy, water concepts - and probably the best application of this tool would be a pre- and post-test. If you would like to do an intervention in your classroom or try a new strategy and see if it works or impacts learning, you could do this test at the beginning of class to see what the baseline is for students coming in because they come in with lots of preconceptions.  Then, conduct your intervention, followed by our test afterwards to see if learning happened, and what kind of learning happened. The fun thing with these machine learning algorithms is, not only do we get just a right or wrong answer, which is what we would already get from the multiple choice type questions, but we also get a lot more nuance about how students connect ideas and where there are potentially barriers to connections based on understanding of terminology or concepts.  Also, identifying where are they struggling to make that jump to a deeper understanding is really what I'm excited about.

What has been the most rewarding part of your research career thus far?

I would say definitely working with students and postdocs, which I think would probably reflect a lot of folks’ experience here at UNC. We're not an R1 university where folks are just doing research and not being teachers.  Being able to work with undergrads, grads, and postdocs has been so rewarding.  It’s really exciting to work with someone, let them be in charge of something and see them make decisions and support them in doing that research. 

You are part of the UNC Million Dollar Club. What is running a million-dollar NSF project like?

The other day I told someone I was going to do some research time and then I said, actually, that just means I'm doing paperwork. There’s definitely a lot of that that goes on, but on a more serious note, there is a lot of project management, which is another thing that I think faculty don't always have a lot of exposure to until we're put in this situation. Before I went to grad school, I worked for about 7 years for an environmental management company where I did manage projects.  I find that I lean on the skills that I developed there, sometimes as much, maybe sometimes more, than other skills I learned in graduate school.

What have been some of the challenges?

My grant started in the middle of COVID, so that was a huge challenge as a lot of the things we had planned, such as travel to the various sites, had to be reimagined.  Obviously, we could not do what we proposed, so the biggest challenge was completely revamping our first year of the grant so that we could do everything remotely.  It did suck up a bunch of time because we had to learn how to use Microsoft Teams, learn how to generate transcripts, and all our changes had to go through IRB.

What advice would you give to a faculty member that is thinking about pursuing research?

Don’t be discouraged.  Lean on other faculty.  Use the support available at UNC.  You can't do the work if you didn't win the money, so you need to get that grant. You need to use all the support you can. I work a lot with ORSP, and I’m messagng people there all the time; like weekly. Sometimes I feel like I talk to you all as much as I talk to my department.  You have Carman assisting with research development, so there's all this support available.  Also I recommend to get your grant proposal under different eyes. For example, I had folks read it who were not in my field at all.  If it doesn't make sense to them, then I worry about whoever is reviewing it being able to understand what I'm talking about.  You can definitely get stuck in your own field with your blinders on and use all these words that don't make sense to anybody else. Getting that proposal in a place where it has a good rationale and flow is so critical for a successful award. But, it's also critical for understanding what you're doing.  This particular grant that I have now, it was the 2nd submission.  We submitted first in 2018, got rejected and then resubmitted in 2019, and the reviewer comments were great. I mean some were very, very critical and that happens, but I carefully considered all of them, and talked to the program officer before resubmitting.

What next for you?

I actually was part of another large grant submission this summer. This one was led by Harmony Newman in Sociology and I'm so excited by the fact that we've got teams on campus of faculty across programs and colleges.  Steve Anderson in Earth Sciences is my current co-PI and he's not even in my same college, and then Harmony is a Sociologist, and we're going to be working together. We're really excited! We submitted to the new racial equity in STEM program. I’ll definitely also be submitting a second NSF IUSE grant like the one I currently have so we can continue that work with the assessment tool once it’s built.