His research focuses on developing methods for causal inference, particularly with applications in environmental health and health outcomes/systems research. He also explores methodological problems involving machine learning and Bayesian methods.
Can you describe a research project where you applied biostatistical methods to address a real-world public health issue?
Like many of my colleagues at the time, I was involved with a study that examined the early health effects from the COVID-19 pandemic. Specifically, we were interested in determining whether wildfire smoke worsened COVID-19 incidence and mortality across the western United States. My colleagues and I collated daily COVID-19 data from nearly every county in California, Washington, and Oregon during the height of both the pandemic and one of the worst wildfire seasons on record, covering nearly 49 million people over a nine-month period.
Our analysis needed to account for the delayed effects of both wildfire exposure as well as COVID-19's incubation period, handle lots of days with zero cases in smaller counties, and control for weather, seasonality, and other confounding factors. We implemented a Bayesian distributed lag model that allowed us to track exposure effects up to four weeks before any incidences and predict what the progression of the pandemic on the West coast might have looked like without the wildfires using historical air quality data. Overall, we estimated nearly 20,000 excessive COVID-19 cases and about 750 excess deaths were directly attributable to wildfire smoke exposure. This work introduced me to environmental health and showed how environmental crises and disasters can compound health effects.
How do you engage students in biostatistics to ensure they grasp complex concepts?
When teaching complex statistical methods, I often confide in my students that I'm working through these concepts alongside them rather than pretending I have immediate mastery. During student meetings, I demonstrate my learning process in real time. For instance, when we look into a particularly dense statistical paper, I'll show them exactly how I approach it. I'll tell them when notation is confusing to me, and we'll try to decipher those equations together. Or I'll break a method apart into more discrete components to try and understand what's happening to hopefully reveal the bigger picture.
This approach demystifies how others, including myself, learn statistics and shows them that confusion is a normal part of the process. I often draw diagrams when formulas become too abstract, and asking the fundamental question of what the method is trying to accomplish before getting lost in the mathematical details. The students become more willing to engage with difficult material when they realize that persistence and systematically approaching a problem with a rigorous routine is more important than immediate comprehension. My hope is that the students start adopting these same strategies naturally, and cultivate the confidence to tackle new challenges on their own.
How do you stay updated with the latest advancements in biostatistics?
I stay current with biostatistical developments through a couple of straightforward but effective approaches. I have Google Scholar alerts set up for a variety of keywords related to my research interests. It's not at all original, but I find it a reliable tool for receiving a steady stream of new papers and preprints as they come out.
I also make it a priority to attend subject specific (e.g. smaller) conferences whenever possible. At these conferences, you get to learn about cutting-edge research directly from those developing new methods or finding new discoveries. Smaller conferences also enable opportunities for informal conversations with other researchers where you learn about practical implementation challenges that never make it into the final papers. Plus, you get a sense of where the field is heading and what problems researchers are struggling with in their day-to-day work. These connections often lead to collaborations or point me towards methods I would have never been aware of through broad literature reviews.
What has been one of the biggest lessons you've learned in your academic or research career?
One of the biggest lessons I've learned is not to become too attached to any single piece of work or to dwell on mistakes I've made along the way. Early in my career, I fell into the trap of thinking that one paper or project would define my entire research career. When things didn't go as planned or when I realized I'd made errors in my approach, I'd spend way too much mental energy beating myself up about it.
What I've come to understand is that research is fundamentally iterative. No single study, no matter how well-executed, tells the complete story. I've learned to view errors and setbacks as insightful rather than personal failings. When I look back at some of my earlier work now, I can identify choices that I would now have made differently. Despite the sometimes-sinking feeling, those revelations and acknowledgements are healthy. It means I'm still learning and growing as a researcher. The researchers I respect the most are the ones who are honest about the limitations of their work and willing to adapt when new evidence emerges. This perspective has made me a better collaborator and much less stressed about the inevitable ups and downs of academic research.
Why did you choose biostatistics (or informatics/data science) as a field, and what keeps you passionate about your work?
I chose biostatistics because it offers the opportunity to be a genuine team member across a diversity of scientific fields while still having a distinct technical contribution. Biostatistics puts you right in the middle of substantive research questions where you are learning about everything from epidemiology to health policy to clinical medicine. I get to work with physicians, public health researchers, and policy experts, which allows me to constantly expand my knowledge beyond statistical methodologies.
What really keeps me passionate is that statistics operates in the gray areas where most consequential scientific discoveries happen. There are very few black and white findings left to be made in biomedical research. Almost everything involves uncertainty, confounding, measurement error, and competing explanations. That's exactly where statistical thinking becomes essential, cutting through the noise to find genuine signals in complex data. Every project presents a new puzzle where I (and my colleagues) figure out the right methodological approaches to get reliable answers from messy, real-world data.
Learn More
To read more about Dr. Josey’s work, visit his faculty profile on the Colorado School of Public Health website.
Stay tuned for more features in our Get to Know Your B&I Faculty series!