People in this episode:
Dr Nicole Beisiegel (performer)
Dr Jessamyn Fairfield (host)
Dr Shane Jensen (guest)
Hi, I’m Nicole and I’m a postdoc in applied maths in UCD. And today I thought I am going to give you a bit of an idea what my research is about. But since Halloween is just around the corner, I also thought I’d talk about something a bit scary. But then I actually noticed for some of you that might actually be the same thing. Honestly, if I would get a euro for every time somebody tells me they’re bad at maths, that combined with my postdoc salary would actually add up to a living wage.
Welcome to You’re Up Next, a podcast by Bright Club Ireland that explores what comedy can do for research and society. I’m Jessamyn Fairfield, and I’m a physicist, comedian, and spreadsheet enthusiast. This is the second season of You’re Up Next, and we’ll be exploring some important topics through the lenses of science and humor. In this episode, we’re talking about math, maths, and statistics. And I’ll be speaking with Shane Jensen about how these can not only be fun, but applied to sports, cities, and the world around us. The power of comedy lies in breaking down, humanizing, and making light of even the heaviest emotional and technical topics. So this week, we’re going to take a look at something that can inspire great fear, loathing, and even dread, as well as great joy and enlightenment. I’m talking of course, about math. Don’t change the channel, I promise this will be funny.
If you feel like there are numbers people and non numbers people, I totally hear you. I didn’t really enjoy math growing up, despite being raised by two people with math degrees. And I only really came around in the last year of my education where math was mandatory. People sometimes talk about math anxiety, where students can get so worried about whether they’re a math person or not, that it occupies their working memory, making them worse at math. But if we set the worrying aside, numbers and numerical methods provide an incredible way to solve problems in the world and make predictions about how it works. It’s a means of abstracting what we see around us, like good versus evil, dark energy versus matter, altruism versus selfishness. Math is a literal, different language we can learn that isn’t better or worse than verbal language, but it lets us do different things. An especially interesting example is statistics, which is rooted in math and data analysis, which has its own arcane ways and means, but which is also a favorite domain of that contemporary villain, The Internet Arguer. Statistics can feel like math applied for the benefit of humanity, and it can also feel like data wizardry that’s being marshaled to support almost any argument.
So this week, we’re chatting to Shane Jensen, who is a statistics professor at the University of Pennsylvania, and who uses and communicates stats are on a lot of interesting societal phenomena.
My name is Shane Jensen and I’m a professor of statistics at the University of Pennsylvania, in the Wharton School here and I’ve been a professor here teaching statistics for I guess 16 years now, which is a little mind blowing. I do a lot of really, what I think, cool research in kind of urban analytics and sports statistics and some other things – those are kind of the areas of application which I teach statistics,
I like to start with a really aggressive question [laughs], so do you think that like math and stats tend to get a bad rap sometimes, as like very technical or boring topics? Obviously not to you because you’ve devoted your life to statistics, but I feel like sometimes people have that impression of them.
Yeah, I get a very what a statistician would call very bimodal reaction basically whenever I introduce myself as a statistician. People like, you know, I mean, I think people either have a relative- I mean most people have a relatively negative I think experience with it through their own education. And I think it’s in part because the way it’s taught is- sucks mostly, I mean. I mean, the way we kind of teach statistics at least at an introductory level, I think because so many statisticians come from kind of more like the mathematical side of things you know, they fell in love in statistics because the mathematical beauty of it or whatever, but that’s not I think what engages, you know, I think most people. In my own experience, teaching from kind of more like in a specific application kind of area, or kind of motivating it more through like real world data ends up being more compelling to people but I don’t think that’s the usual style in my field. So yeah, I mean I’m not surprised that people have that kind of reaction. I’ve been kind of fighting it my whole life, I suppose.
The one thing that has been kind of nice, over at least the last couple, you know, decade or so is I no longer- people still say “I had a terrible class in that” or “I had a terrible experience trying to learn that”, but they no longer say “why would you want to do that?”. So the nice thing is is statistics and you know, kind of analytics and machine learning or whatever else we want to call it has become so kind of commonplace in society, people no longer, no longer kind of question why people are doing that at least. So that’s one battle, I guess, that I no longer have to fight
That is definitely a positive result. And because yeah, I know that you’ve taught a very broad variety of students, and you were mentioning too- like, I know, being in the Wharton School, you teach some MBA students who maybe do not want to learn statistics as much as other students. I don’t know if I’m making a generalization here.
I mean, that is certainly the- yeah, I mean, they they are a little bit, you know, in some ways, they’re, you know, they certainly come across to me, at least as the, you know, maybe the ones that are the least kind of, you know, into the inner workings, the ‘under the hood’ kind of thing, but at the same time, the MBA students, they were the ones that really kind of, at least first push me to kind of think about this kind of taking things from more applied perspective, because they are certainly I mean, almost to a fault, constantly questioning kind of what is the real world? What is the tangible value of each of these kind of procedures or methods that we’re learning? I mean, they again have their own unique perspective, like, “how will this make me money, you know, very immediately?”.
[laughs] Yeah, well, and I mean, like, let’s talk about some of those applications, because I always find it really interesting, the sports statistics stuff that you do. And I don’t know if that’s the kind of thing that you bring up in your teaching, or I know you’ve done a lot of kind of public facing outreach about it as well, because people are interested in sports.
Yeah, no, and I mean, like, in some ways, it’s kind of a bit of a risky kind of application area to try and engage students with because, you know, people typically are either really into sports or really not into sports. But I really, I mean, I personally am into sports. So I mean, you know, I certainly can teach enthusiastically using it, which is nice. And I feel like even you know, I manage to I think engage a lot of people that aren’t into sports, because I try and kind of teach a law that kind of applications almost like from a personal level and I think you know, you know, that plus you know, a lot of kind of self effacing humor is really kind of helpful in that direction. Like for example, you know, I did this big project involving involving fielding in baseball many, many years ago. And, you know, I also kind of got into baseball specifically because I was in grad school in Boston, and it’s almost a religion up in that area that you’re into baseball. And you know, I’ve taught mostly on the east coast where you know, everybody’s kind of conscious at least of this big rivalry that exists between New York and Boston in baseball, you know, the word the fans don’t like each other, the players don’t like each other. It’s been one of the big rivalries in American sports over the last 100 years. And I got to actually kind of engage that directly because there are some prominent kind of players in baseball that I got to evaluate as part of this project. And one in particular that even even non fans of baseball know is this guy, Derek Jeter, who used to play for the Yankees. He in fact is a very famous player – first ballot Hall of Famer – and we got to basically analyze his fielding. And you know, again, in Boston, there was this constant sort of mantra that Derek Jeter was overrated, and all this type of stuff. And I got to kind of basically dig into the data and at least you know, according to our methods, you know, Derek Jeter – though he’s an amazing hitter, actually came out as kind of a poor fielder, you know, according to our methods. And that’s something that Bostonians just ate up. In fact, I remember giving a talk at this in a conference in Boston and it was one of the big American Association for Advancement of Science, so it was a rather big meeting and so there was was some press there, and I got you know, interviewed by this Boston Globe person that did a kind of write up of it. And it was so funny to me to sort of see that then you know, it was picked up by the Boston Globe and then the New York Post wrote an article that was- yeah I mean, it couldn’t have been more more opposite in terms of its take on my research.
Yeah, it’s funny too to imagine like a research talk that you could give that could get like, you know, a standing ovation in Boston and like booed off the stage in New York City just like “no don’t say that”.
And it was delightful too because the New York Post- I mean again the New York Post in general does not have a great reputation because it’s kind of you know, it is kind of you know, not the most intellectual paper out there, but they always have that kind of whole like “now let’s go to the man on the street” kind of thing, and they did that for this particular article and it’s like you know, they talked to like Joe in Queens he’s like “what are the professors down there at Penn smoking? They don’t know anything. Jeter’s got a lot of heart.” So it was really kind of funny to see. You know, I’ve kind of grown up reading things like the New York Post and being like “oh man, that’s not a very- you know, that’s that’s a pretty superficial take”, and then I got to have the New York Post take on my own stuff.
And you talked about, like humor there, as well as like, both useful in terms of like reaching students and people that might not be that inherently interested in math and stats. But I assume it’s also helpful in terms of like communicating this important Derek Jeter related research to the public.
Yeah, no, and especially because nowadays it’s gotten again much easier as kind of stats and machine learning and all these things have kind of, you know, pervaded our lives basically. So these days, when I teach something, you know, I mean, I might teach a topic like, you know, regression, you know, or just predictive modeling in general, and, you know, used to be more difficult. Now, I can just kind of point to being like, you know, this is the kind of model that, you know, Amazon is using to make its recommendations to you or Netflix is using to make its recommendations to you. And I usually kind of present that in some kind of tried like, humoristic, like, self effacing kind of way, where I kind of talk about basically what Amazon does, with my recommendations, and how, you know, through predictive modeling over the last couple years, it’s learned to throw, like, every single piece of clothing that has a wolf on it at me. And you know, I keep buying it, and I kind of can kind of talk about almost like reinforcement learning, and all these relatively sophisticated concepts, you know, with this tangible example of like, how I started buying t shirts with wolves on them, and then it started recommending more t shirts with wolves on them, and I started buying those. And, you know, it will occasionally try and throw it a t-shirt with a dragon on it and see if I like that too. And through that it can learn that dragons are correlated with wolves, and I can kind of start talking about how classification algorithms kind of work in general.
Yeah, well, and because you did a radio show for quite a while about sports statistics as well, right? So like, what were the kind of communication tools that were…
What I like about about the sports podcast is we are kind of looking at sports from a more analytical perspective, and it allows me to kind of keep up with the field. But we also kind of, you know, do approach it with, I think humor, but I think also humility. Because I think a lot of especially academic takes on things – and I fall into this trap myself, especially from a statistical point of view – it’s so easy to look at a study out there that’s maybe been done, and sort of start poking holes in it immediately. Because every single analysis, no matter how principled, you try and be, you know, there’s subjective decision making involved and, you know, hopefully that’s informed by science or whatever, but you’re going to be making subjective decisions and judgments all along the way. And then so it’s always easy to retrospectively be like, “oh, I wouldn’t have done the exact same thing in that situation”. But I always kind of, I think, try and do a good job on, you know, when we kind of look at what studies are out there of trying to kind of not focus entirely on the negative, because I think you can quickly kind of get lost in the weeds with that. Myself and my co hosts, we always have it as kind of a conscious goal to sort of, like, you know, emphasize more the big ideas that are coming across this paper, even if we kind of, you know, maybe this technical issue wasn’t done well, or this technical issue wasn’t done well. And when you kind of have that perspective, you can kind of realize how much amazing research is going on out there in sports analysis, for example, and a lot of it’s not being done necessarily within the academic community, it’s being done by teams, or, you know, just kind of, you know, people blogging out there and stuff like that.
Yeah, well, I think that’s such a great point, too, that like, you know, it’s an application of statistics that has very broad interest beyond academia. And I think too like, what you said there about the, you know, the fact that choices are being made in how these things are analyzed, like, I think there can be a temptation of people to say, like, “oh, like, it’s just, it’s just the numbers, like, the numbers are the numbers man” or something like that. When it’s like, well, no, we’re making choices in how we analyze these things, and that’s going to affect results. And that’s actually a really important process to understand, and that doesn’t necessarily get communicated that much.
And even once the numbers are analyzed, how you interpret said number. I mean, you know, it’s really easy to come up with examples these days, especially like in the geopolitical world we live in where like, you know, the same statistic, the same number can be completely interpreted completely differently, depending on the prior viewpoints or biases, or whatever you’re bringing into things. And so I think, you know, certainly within the US with the last couple of election cycles, and the amount of information and misinformation that’s been going on out there, and how that misinformation is becoming like all information more quantitative in nature. You know, it’s been both kind of exhilarating, but also incredibly frustrating as a statistician to sort of see that happen because there is so much misinformation out there where, you know, a particular number will be presented without the necessary context to understand it. And so it’s kind of easy, certainly from the teaching side, to kind of motivate how important it is to be thoughtful about your interpretation of quantitative data.
I have to imagine too that in terms of like using the same number to support multiple interpretations, like do you want to tell us about your your current work on urban analytics, because I have to imagine that that happens a lot there.
I’ve kind of gotten really into over the last like, five, six years into studying cities in general, but Philadelphia specifically. And so like there’s a lot more kind of data out there to kind of analyze basically, how cities function.
Have there been any surprises in that kind of research, stuff that might have gone against the conventional wisdom around cities or urban planning?
I mean, I guess I shouldn’t have expected it would have been an easy endeavor anyway, because cities are very complex, interconnected things, but you know, just the amount of nuance there is to resolve. You know, like, there’s a big kind of initiative in Philadelphia to take vacant lots and turn them into little parks or pop up gardens and stuff like that. And, you know, of course, the hypothesis would be that that’s going to be good for the area. And we do know, overall, we do see things like improvements in safety and stuff like that, around these kind of little parks versus kind of, you know, the control group of lots that remain vacant. But it’s subtle, these are relatively small effects and on kind of like a yearly scale, but like, you know, on a decade scale can be quite substantial. So you have to kind of look at things at the right timeframe. But also, you know, there’s a lot of kind of surrounding context that influences that. You know, a new park is going to be more effective in terms of promoting public vibrancy if there’s already a café nearby that people can kind of use as part of that. And so that type of stuff, I think, is, you know, where we’ve kind of- you know, we were a little bit surprised the extent to which the surrounding context does matter. You know, again, planners can be a little bit more thoughtful about kind of taking into account the surrounding neighborhood context when they do these interventions.
I’m not an academic, that just sort of thinks that it’s easy to kind of just do these studies, and do the analysis. I kind of realized how difficult it is to basically get any kind of intervention, you know, actually into a city, and how complex things are just in general. I think, you know, I use a lot of humor, kind of in that connection point. And I think that also helps, too, because, in my experience, when people meet a new academic, you know, like it’s a government public policy maker, or just, you know, somebody at the city’s Office of Information Technology, I think they’re maybe a little bit on the defensive, because academics can be kind of surly, and can be kind of, you know, high and mighty about their knowledge. That first connection between somebody and an academic often doesn’t go as well, because academics tend to come in and think they know everything. And so I think, again, that’s sort of like, you know, the nature of my humor, which is, you know, mostly a long string of admitting what I don’t know about things, I think somehow it does actually help.
Admitting what you don’t know is a great start in making any technical topic more accessible. I also think it’s really interesting that humor can be so useful in defusing anxiety around math and other things, and busting those imagined hierarchies of people who get it and people who don’t. If you Google math jokes, you will find a lot of in-jokes that don’t really communicate math at all. They effectively just congratulate an in-crowd that already knows something. But what Shane described of making the surprising applications of statistics clear, while also being self deprecating and listening, that’s the kind of STEM humor that I’m here for.
I mentioned at the start of the episode that I didn’t really like math until I was a teenager. And if you’re wondering what turned it around for me, it was an amazing teacher I had: Joy Hansberry. She was fun, laid back, but also really good at explaining things and connecting to a wide variety of students. She let students bring in their own music to work on problems to. She give high marks for a project about integrals that I turned in with math jokes on every cross section. She had a velvet painting of Elvis at the back of the classroom. And she taught the most basic and the most advanced math classes in our high school, and was beloved by both groups of students. A great teacher can draw students to any subject, but to me what was really amazing about Ms Hansberry was that she took this subject that people thought of as very arcane and scary, and made it fascinating and accessible to anyone. And, yes, even funny. So this episode is dedicated to her.
This episode was made possible by support from the community knowledge initiative, the CÚRAM Medical Device Research Center, the research office at NUI Galway, and Science Foundation Ireland. We’re grateful to our guests, our host, which is me, Jessamyn Fairfield, our producers Maurice and Shaun, and to you our listeners and our Bright Club Ireland community. A transcript of this episode is available at brightclub.ie, where you can also find more information about what we do at Bright Club Ireland.