Some time ago, I interviewed Dr. Liz Bradley from the University of Colorado Boulder for a story in Nature Methods about athlete-scientists. It's called Athlete-scientists like to sweat and is a column in the series Lab&Life.
Here is a podcast with more from that interview. Dr. Bradley is a mathematician and a rower. So you will hear about sweeping and sculling, about rugby, about math, about the importance of notebooks and why whitespace can help you and your science.
Action shot of Liz Bradley at the 1988 Olympic Games. Her team placed fifth in the women's coxed four event. That's a boat with four rowers with a cox, or coxswain, who is in charge of steering and navigation, safety and motivation.
Action shot of Liz Bradley with PhD student Abby Jacobs and Aaron Clauset, a colleague in the computer science department at the University of Colorado Boulder.
Part 1 in this series is a conversation with runner and climber Dr. Gene Yeo, an RNA biologist at the University of California San Diego.
You can listen here or on the streaming services of your choice. It's a series called Conversations with scientists. On Apple podcasts, it's here. On Google podcasts, it's here. On Spotify, it's here.
Transcript of the podcast
Note: These podcasts are produced to be heard. If you can, please tune in. Transcripts are generated using speech recognition software and there’s a human editor. But a transcript may contain errors. Please check the corresponding audio before quoting.
When I was an undergraduate, I was on the rowing team. And that was a lot less hours and a lot less intensity than when I was a grad student and I was on the national team. Because the, you know, the college team trains once a day, for an hour and a half. National team, you're training twice a day for a couple hours, at least each time.
That's Dr. Liz Bradley who is on the computer science faculty at the University of Colorado Boulder. Hi and Welcome to conversations with scientists, I'm Vivien Marx.
Some time ago, I interviewed Liz Bradley for a story in Nature Methods about athlete-scientists and this is a podcast with more from that interview. Dr. Bradley is a mathematician and a rower. So you will hear about sweeping and sculling in this podcast. She will talk about rugby, too.
You will hear about her approach to thinking and about time management, about why it's good to have a notebook, about data analysis and some pitfalls that can happen in data analysis and how to prevent those and why it's a good idea to not just use tools without understanding what they do.
There is a Wikipedia page about Liz Bradley and the link is in the show notes.
Liz Bradley [1:15]
I didn't even know it existed. I'm not involved with it. But yeah, it looked mostly accurate.
She competed in the World Rowing Championships in 1986 and 1987, finishing fourth and fifth. At the 1988 Olympic Games, her team placed fifth in the women's coxed four event. That's a boat with four rowers who each have an oar and there is a cox or coxswain who is in charge of steering and navigation, safety and motivation. The Wikipedia page describes her as a mathematician and an Olympic rower.
Liz Bradley [1:50]
We all have lots of identities. And, you know, people are mothers, fathers, sisters, they identify by their sport, they're identified by their, you know, gender preference orientation. You know, and there's lots of attributes, but the I guess the most important thing to me about myself is what I'm interested in intellectually. That's not completely true. Because relationship is important.
And performing arts are important. And, and physicality is, is essential. It's not important. It just has to be there. So you asked if I train anymore, I don't, I'm 61 I've been skiing for 55 years, my knees are destroyed. Well, one of my knees is destroyed. It's going to be metal next year, which is hopefully going to be better. I played rugby, which was how the knee got hurt.
I was actually on the New England Select side in 1979, maybe something like that. So I was an I actually had been to Nationals in rugby. But that I got thrown off the rugby team in 1982 or three because I got my third concussion. And then I started rowing. So I started rowing quite late.
She was thrown off the rugby team. And I should add it was because of her concussions.
Liz Bradley [3:10]
The medical department at MIT said you've gotten a third concussion, you should not play rugby anymore. No, rugby is actually less dangerous than American football because you are not padded. A lot of the injuries that happen in American football are because people are trying to kill each other and they've got pads on. Rugby. If you tackle somebody hard, it's going to hurt you as much as them. So you tend to use technique instead of just brute force.
If you learn to put your head behind the person's legs rather than in front of them if they're running when you're tackling them, you know that football players don't learn this stuff because they don't have to.
In spite of her injuries, she just loved the game and still does.
Liz Bradley [4:15]
Rugby is so much fun. And also, you know, I'm six foot one. And I'm not fast over long distances, but I was reasonably fast over short distances. So rugby was really fun.
She doesn't play rugby anymore but she does plenty of sports activities even though her injuries have limited some of that. And she has advised on athletics at the University of Colorado.
Athletics is not so much hard as I've had to adapt. So I still ski probably 40 days a year. I still bike pretty much every day. But I can't run. I can't hike. You know, there are things I can't do. So I've had to adapt. I've moved to Telemark skis from downhill skis, which helps your knees a little bit.
I was involved in athletics oversight at the University of Colorado, which was quite interesting. I was a full, Professor tenured, and I was an Olympic athlete, so I could kind of talk to both sides. And so I was like Madeleine Albright in the middle of, you know, the Middle East with these two warring factions. And trying to bring them to the same place was interesting. So I did indirectly work with student athletes, but mostly, by way of helping the administration and the athletic department. Come to the same point about, you know, academics, you know, what are the requirements? And how are we going to support these people? And no, we don't want to have tutors that just write their papers for them. You know, things like that.
I asked what impact athletics has on an individual student's life and how she prepares people for making the decision to do both school and athletics.
Liz Bradley [5:35]
I certainly do encourage it, I do warn people that they are going to have to do a lot of very careful prioritization and task juggling and, and contact switching, as we've seen in computer science.
And, you know, then this is you lose some spontaneity, because you're always really having to decide which what you're going to do with which half hour hour a day, and that that's not good in the long term for a person because you know, if you think about any of the Eastern religions, they tell you that there's no and even Thoreau, said, I like a broad march into my life. I know this business of the whitespace in our lives is part of what makes us creative.
And what makes you know, this is there's a reason we have good ideas in the shower. y6Because that's where you're not trying, you're just kind of letting things to in your head. So if you're in a sport, where like, I suppose, long distance swimming, you can probably do a lot of thinking during long distance swimming, but rowing you can't because it's a constant battle against you know, trying to keep your technique in the face of basically pain.
Because you're you go off the starting line, as hard as you can. And within 30 seconds, your heart rate is pegged, and you're just basically trying to stay alive for the next six or seven minutes and not you know, while making the boat go really fast. So you're always titrating you're, you feel like the red coming in and you back off just a little bit till it goes out. And you know, your whole body's screaming. At that level. It looks when a rowing shell that's moving really fast looks effortless, but it's not.
Hard to know what it feels like to row I can only imagine what it feels like to have your whole body screaming as you practice a sport with that intensity.
With rowing you can row using mainly one hand or two. If you use two hands to row you are a sculler and with one oar you are a sweeper.
Sweep and scull.
You did both, I think, right?
Yeah. But I'm mostly a sweep rower. So when you row, you pick a side, and you row on that side, and you very rarely switch. In fact, if you switch, you tend to get hurt. Because there's a lot of asymmetry and a lot of infrastructure that your body forms. And then if you switch sides, you still know how to pull hard, but your body doesn't have the infrastructure. And so well, to begin with, you know, your hands get destroyed, because you have calluses, you know. This hand is just a hook, this hand actually turns the oar. So it's different set of calluses, different muscles, I still have one bigger pec than the other. And it's been thirty, God knows how many years and one bigger leg.
I asked her about the traits that are needed in sports and to what degree they help in science, too. Rowing, she says, is absolutely and intimately a team sport. More so than any of the half dozen other sports that she has played seriously. What matters in sports and in science are things like persistence, stamina and discipline.
Liz Bradley [8:40]
Discipline, you know, you're, you can, if you can, sport teaches you how to focus, and how to concentrate in ways that involve your physical body as well as your brain. The body and the brain are not separate things, they talk to each other. And if you're, if you train the whole organism to concentrate, then the brain is going to be carried along as well. I think that's important.
I wondered how, when Liz Bradley was a student how she managed to fit school and athletics into her day. As a former fellow at MIT, I know a bit about MIT and busy schedules from when I was there. But I remember student schedules were much busier and more intense than mine.
She was on the rowing team as an undergraduate at MIT and then as a graduate student at MIT she was on the National Rowing Team. With that athletic background and as a scientist, I wondered how she advises others who are thinking about combining school and athletics.
When we were talking to people about whether or not they should join the rowing team, they would often say, but I won't have any time for my homework. And we would say, and it was true, that when you have to structure your time, and you know, you're gonna go to practice, you don't kind of screw around and say, Oh, well, I'll do my homework later. Because you know, you've got to do it now. So it's makes you better at time management, typically. In the same way that deadlines make us sometimes more effective.
Now, that being said, when I was an undergraduate, I was on the rowing team. And that was a lot less hours and a lot less intensity than when I was a grad student and I was on the national team. Because the, you know, the college team trains once a day, for an hour and a half. National team, you're training twice a day for a couple hours, at least each time. It's not like swimming, where the workouts are, you know, hugely long, because rowing is intense. And you can't do that for more than a couple hours, which is nice.
But you know, you come back from from practice, and I would fall asleep in the beanbag chair in my office for an hour or so. And then I would start working on my research. My advisor was incredibly understanding, which was the only reason I was able to do this.
She had two advisors Hal Abelson and Gerald Sussman.
Liz Bradley [11:00]
Both of them were very, my advisors were very understanding. |Hal and Gerri kind of tag team-advised most of their students, so people will say they're Hal's or Gerry's, I was more Gerry's student than Hal's, which is why I say Gerry.
She takes the way she was mentored into her own mentoring with her students and trainees.
Yeah, I do. And, you know, I have to be a little careful, because I'm not their mom, and I'm not their shrink. But I do talk to them about, you know, the whitespace. And making sure that you don't spend your entire day trying to have an idea. But too, and I actually, I really require all my students to have a physical, like, physical hardcopy notebook. And they, they look at me like owls because they're used to doing everything on a computer. And so I give them one I buy really nice notebooks. This is not for my class students. This is for my research group. And one of the first meetings I have with them, I give them this physical notebook, and I show them how I use mine. And I say, there's different ways to use it. But it's different, you know, again, we, we evolved, there's proprioception, there's the fact that you know, you can, you can draw, and yes, tablets let you do this somewhat, but you know, having a physical notebook, it's all there. And I will advise them sometimes, if they're stuck, to go to a coffee house with their notebook and a pen, and no computer, and think and write.
And so you've probably, you know, this is echoes of, what's her name, the artist's way, lady, you know, the the writing, you know, just writing.
One guy, this is actually really funny, he finished his thesis proposal. And he said, What should I do next? And I said, we'll put it aside for a week and then print it out and take it to the Trident coffee house with a pen, don't take your computer, he looked at me like I had three heads. And then he came back about a week later, kind of like he'd seen God. He said, my God, you know, I was able to read it in a holistic way and see the whole thing and I was able to write on it. And I could remember things by where they were on the page, and it's so much easier to flip back and forth.
And I didn't know whether to be delighted or cry that this is the first time this young man had ever edited his writing on paper. Ever. And, you know, some of us are old enough to have had that affordance. But yeah, so I do work with my students on not so much how to think but, but some of the strategies that I've developed to to make sure that there's that whitespace, that there's that margin that there's that place.
I took a modern dance class once and had wonderful teacher he said, open the head, put in this concept, close the head. Don't think about anything, just dance. So it's not about trying. It's about just letting things evolve organically.
Empty the head, whitespace and a notebook. Those are part of her strategy to work with the members of her lab. The notebook, by the way, is not a lab notebook and not anything she asks to look at. But she will show her notebook also to explain how she uses it.
This notebook is not a lab notebook is something you want to see]
Oh no, it's totally private. It's nice. Like, sometimes I'll take notes on something like this was from a meeting with a student, there's some drawings, and there's some text and questions and questions, complex unmodelable stuff on the right hand side, you know. And I'll show that to the student. I'll say this is, you know, this is what I was looking at the stuff that struck me from our meeting. But no, I don't look at their notebook.
The notebook is part of her approach to mentoring.
If you get them a really nice notebook, it seems to make a difference. Because there's this beautiful artifact. And you know, my graduation present to all of them is a really nice pen.
Next I asked Liz Bradley more about her science, because she is involved in biology research projects and collaborates with colleagues at the Santa Fe Institute. She also talked about how applied mathematics powers her data science.
Liz Bradley [15:35]
So actually, I don't have a degree in computer science, my degree is in electrical engineering, what I really do is, I guess what would be called Applied Mathematics. So I am not really interested in algorithms. I'm not even really interested in code.
I'm interested in analyzing data. And a lot of the tools that I use to analyze data are just way too onerous to do with paper and pencil, so we get the computers to do them, that's all.
She is involved with BioFrontiers, which appears to me to be institute where scientists who work in different disciplines are all under one roof and where there is discussion space for them to interact in all sorts of ways.
BioFrontiers I'm affiliated with only because of administrative history, I was the department chair when Bio-Frontiers was being formed. So the University of Colorado has these really cool things called institutes. And they are designed to break down the barriers between departments. And so there's an institute for Alpine Arctic research, there's a cooperative Institute for the Research of Earth science or something like there's a bunch of these things, JILA, Joint Institute of Laboratory Astrophysics, which is the home for three of our four Nobel Prize winners. So if you have a field that is interdisciplinary, and physicists work in it, and chemists work in it, and biologists work in it, and maybe a social scientist works in it, it's very, very hard and traditional academic setting to do that.
And so our Institutes at the University of Colorado are designed to do that. Bio-Frontiers was forming when I was chair. And it was pretty clear that computer science shouldn't be included. So I got involved to basically put my shoulder against my department and say, You need to be involved in this. And so I've stayed involved. Because when you do something in academia, the roots can be you need to you need to stay involved to the roots get pretty deep or else the thing will get pulled out and go away.
You don't want a something you have developed to go away, I found that an intriguing bit of insight about academia. I asked her some more about her own research.
Liz Bradley [17.45]
Basically, I guess I'm a data scientist. That's the new term for it nowadays. So I'm interested in a particular kind of data that's very hard to analyze. And I have various kinds of mathematical techniques that I applied to the data. Where I came from was what's called Chaos Theory. You know, so I did a lot of work on, on analyzing and controlling chaotic systems. I'm involved with the Santa Fe Institute, and I have been for God 25,30 years, because it's a wonderful think tank where people think about, you know, chaos and complexity, and so on and so forth.
She has some favorite bugbears, things that bug her that she sees others do in their data analysis. One of those issues is when scientists use data analysis tools blindly, without thought. And the other is when people don't report how they processed their data and only reporting the 'cooked' data not the raw stuff.
So I'll give you an example. I work with scientists who drill cores, ice cores out of an Antarctic ice sheets mostly. And they wanted to understand the spectral content of the signal. So basically, an ice core records is a is a time history of the climate of the Earth. And so there's ice from 30,000 years ago, it captures what the air was at that time.
So they want to know something about the periodicity is: Were there any sort of cycles in them. So they downloaded a piece of code that was supposed to do that. And they ran it. And they didn't realize that that piece of code, the method that's embedded in that piece of code, has a bunch of different knobs that you use to control it. And when you write code like that, you have to pick a default value for the knob 3,17.
And those default values might or might not be the right thing for each new problem. But everybody just uses the defaults. So yeah, so that's the first bit which is, you know, using these things blindly. I said, What do you How did you do this, or we just use Case Vector or whatever it was. So I downloaded Case Vector , and I got the I took the data, and I changed the knobs a little bit, and the results completely changed. And so they rapidly backpedaled from reporting those results.
That's a painful lesson related to dangers of just using default values with a specific software tool.
Liz Bradley [20:15]
Another instance was, so I work with these people. So I'm able to get the completely raw, unprocessed data. That's not what's put up on the website for public use.
What's put up on the website for public use is processed, and it's processed because oh, that's dirty that, you know, that one was clearly above that, oh, there's a missing number. Let's just take that, you know, all the things that.
Not just outliers, missing values, all sorts of stuff. And so people want, you know, everybody wants their data to be beautiful, and they'll snark at other people if their data is not beautiful. So everybody cleans up their data and they filter it, and they do this and they do that. F iltering takes out meaningful content. And people don't know that. So we were able to, by doing an analysis of the raw data, get information out of it, that fed into a published paper that was not available in the processed data.
It's valuable to do data analysis with an awareness of what the mathematical process is that a given software tool applies.
Another aspect she finds troublesome is when scientists try to reduce the complexity of their data but not exactly the right way, even if they use, say, an approach such as principal component analysis, PCA.
PCA is a particularly good way to lose content, principal component analysis. There's a lot of motivations, some of which are pretty nefarious.
The ice core world is pretty competitive. And so people don't want to publish their good data until they've squeezed everything out of it they can. So they publish dumbed down data. Some of it is just, you know, honest ignorance. They don't know that the processing steps that they're using are obfuscating their hypotheses. They just don't know. And so you know, some famous person wrote a paper that said, you should use a value of three for this particular tool, and then the entire field uses three forever after.
I wondered how this plays out also for more junior scientists, PhD students or postdocs who might be putting themselves under pressure to move forward on a project . They might not feel like they have the time to get fully acquainted with a tool they want to use.
Liz Bradley [22:40]
The tack that I've taken is, I've gone straight to the top, I've gone up to the National Academies, National Academies of Science and Engineering. And I have a colleague, I know people there because of some other work that I've done. So I emailed him. And he got me invited to a workshop where I ranted about this. But no, the way, the way to get these people to understand that is to do what I did, which is to say:
Okay, I'm going to repeat your analysis, but I'm going to change the value. I'm going to slightly change the analysis in a way that, yeah, so basically turn one of the knobs that you didn't bother to see how it should be set and your hypothesis is gone.
That usually gets people's attention. But no, I mean, you can't, you know, and it wouldn't be fair wouldn't be right, to put everybody in my field in service of every scientist in the world to fix their problems. So no, this has to come down from the top. It has to come from National Academies. It has to be a thought piece in Nature or Science. And I'm actually kind of trying to work on that.
Liz Bradley has many projects to tend to tend to, she works across disciplines, she mentors her trainees. She does her own research Here is what she has to say about the life of a professor.
Liz Bradley [22:55]
The life of a professor is how do I say this? We choose our work because we love it. But that's both a bug and a feature. Because if you love it and you feel responsible, and you have students and all that, you need to be careful not to put too much onto your plate.
That was Conversations with scientists. Today's episode was with Dr. Liz Bradley, mathematician at the University of Colorado, Boulder and former Olympic rower.
The music used in this podcast is Freerolling by T Bless and the Professionals, licensed from artlist.io. And I just wanted to say because there's confusion about these things sometimes. Nobody paid for this podcast and nobody paid to be in this podcast. This is independent journalism that I produce in my living room. I'm Vivien Marx. Thanks for listening.