Mike Baiocchi: Stats for Better World Health
April 25, 2019
For Mike Baiocchi, an assistant professor at Stanford Prevention Research Center at Stanford University, statistics is an important part of improving health care. Mike studies causal inference and uses statistics to determine if certain humanitarian intervention efforts, like reducing rates of sexual abuse, are effective in communities. Much of his work is focused on improving the lives of women in sub-Saharan Africa.
We talked with Mike to learn more about how he uses statistics in his day–to–day life and how he first became interested in statistics.
How did you first become interested in statistics?
I became interested in statistics in high school, but I didn’t know it was statistics at that point. I had a lot of fantastic teachers at my high school who were asking me the kinds of questions that are fundamentally statistics and it took me a long time to realize that that’s actually where I was heading.
What does your research involve?
A lot of my research is about prevention, behavioral interventions that are designed to prevent things. That’s really difficult because you don’t get to see the outcome. You don’t get to see that something didn’t happen.
I have to use statistics almost daily. I do that in two ways: one is figuring out very careful measurements of what’s going on, but then I also have to spend time building new techniques because these are really complicated questions in really complicated environments. So being a statistician allows me to both know how to analyze the data but then also construct and put the data together in special ways that haven’t been done before.
What’s an example of how you’ve used statistics in the field?
I use statistics in a few different ways, one is to take complicated data and put it together in ways that people can understand, but another way that being a statistician helps me do my work is we work on behavioral prevention programs, and we have this program where we’ve been teaching people in the slums of Kenya how to improve their situation, how to intervene and stop stuff that they don’t want to have happen.
[Our research includes] a really cool program that the girls [we work with] really enjoy. And one of the issues [during our research] is that they start training each other, they start training their mothers and sisters – and that’s great but eventually, it’ll diffuse. And the treatment group, which is getting a lot of benefit, is all of a sudden leaking over to the control group and the control group starts inching up, so if we’re not careful we lose the ability to detect what should be the whole difference.
So my team, we’ve been investigating ways to deal with this interference problem, with this spill over problem so we can get the full credit of this behavioral intervention.
Why are statisticians crucial to what you do?
It gets really confusing really fast, and we’re the guides. We can really help the rest of the applied folks figure out the right kinds of questions to ask.
What non-statistical skills are important to have as a statistician?
There are a couple of different skills you really want to have in addition to your stats skills. I would say that understanding the content area really well will make your life easier because you have to communicate your ideas out. We can’t just be isolated in our little ivory tower and have really good math, we have to be able to communicate this in a way to the applied folks that they’re going to care about what we’re talking about. So taking those English classes matters, taking a philosophy class matters, and understanding some of the core issues of what’s going on that’ll really make you a much more effective statistician.
Why is statistics important for everyone?
A statistics course is really useful for folks who aren’t even going to spend a lot of time in mathematics or even science because there are a couple of key insights we have as statisticians about how the world will kind of trick you. If you’re not very careful you can have incorrect conclusions. The data, the way things will look is not actually the way the real world is, and if you think carefully through what we’re doing, that’s what we’re trying to help people with.
Don’t get fooled by chance variation, don’t get fooled by this, don’t get fooled by that. And if you spend some time really listening to statistics, and really listening to what it’s trying to say, it’s not just mathematical formulas, there’s something deeper there about actually understanding the world and engaging with it directly.
Meet More Real Statisticians:
- Leslie McClure: Careers in Biostatistics Research
- Effecting Change with Statistics and Data: An Interview with Statistician Talithia Williams
- World AIDS Day: Statisticians Making a Difference
In this year’s Fall Data Challenge, 55 teams submitted their recommendations on how to help solve homelessness in Los Angeles, Seattle or New York City using the U.S. Department of Housing and Urban Development’s 2018 Point-in-Time Estimate of Homelessness in the U.S. dataset. According to HUD, an average of 553,000 people in America experienced homelessness on…
Statisticians make important contributions to many fields, and one where they make a huge impact is in cancer research. We asked Dr. Sujata Patil, Associate Attending Biostatistician at Memorial Sloan Kettering Cancer Center, and Ruth Pfeiffer, Senior Investigator at the National Cancer Institute, about their research, the impact their work has and why they choose…