SPEAKER_01 0:00

In a 2018 study that was done, the researchers determined that 90% of the world's data at that time, this was four years ago, 90% of the data in the world had been created in the last two years. Welcome to Room to Grow. I'm Curtis Brown. And I'm Joni Fenderberg. We work together at Texas Instruments, and we're glad you're here.

SPEAKER_00 0:22

We're looking forward to continually improving our practice, and we understand that you are too.

SPEAKER_01 0:26

We hope that you'll find this podcast as a room for you to grow along with us as we wrestle with and explore ideas about teaching math even better. In today's episode, Curtis and I discuss ideas around data science, an important topic in many states across the country right now. We begin by defining data science as an interdisciplinary, complex topic where math, computer science, statistics, and specific content expertise come together to extract knowledge and insights from noisy, unstructured, and structured data and apply that knowledge across a broad range of domains. We consider why data science is important for math educators and acknowledge some of the criticism in the field against data science in schools. Finally, we wrap up with some ideas and resources for learning more and getting started with data science in the classroom. There's lots here, so let's get growing.

SPEAKER_00 1:20

Well, Joni, it's really great to be chatting with you once again. We have an opportunity to record another podcast. Uh, for me, it is just the very beginning of um the raining season. We've started getting a little bit of rain in Dallas uh this last week, which is fantastic because we've been dry all summer long. So I'm really excited about that. Um, but Joni, we're gonna be talking uh about um data science today. And I'm really excited about this uh this topic and this new uh thing that's going on in the math world. So, Joni, how are you doing?

SPEAKER_01 1:56

I'm doing great, Curtis. It's it's really great to be back together recording again. Um, I'm looking forward to being able to see you in person, but um always good to have these conversations. And yeah, data science has been on our list of topics we want to record a podcast episode about for quite some time. I I think there's a lot of really interesting information here, and it's a really intriguing topic for uh the educators that are that are um our listeners. So yeah, let's dive into the conversation. And I think the best place to start is let's talk about what is data science anyway. I think, you know, in in my exposure to this idea of data science over the last couple of years in the math education space, it's been I've heard data science, I've heard data literacy, I've heard, oh, it's just statistics, or oh, this isn't really math. It's a career field, it's a job title. Like it's it's kind of confusing language, right? So why don't we start by talking about how what we think data science is and how we're talking about it for today's episode.

SPEAKER_00 2:55

Yeah, for sure. I think that's a really important place for us to start uh talking about, just to have a level playing field and make sure we're all on the same page as we're having this conversation. So um, you know, I took a couple of uh data science courses this summer. I had an opportunity to go um and take and study a little bit uh about what other folks were calling data science. Awesome. Uh so that was a that was a really great uh exposure for me because I think I had a a an idea of what I thought data science was when I went into those courses. And now I think I have an entirely different um, you know, idea of what I think data science is now. And whenever I say the words data science, um, so here's what I here's what I uh envisioned. Um I am uh thinking about data science. Whenever I I say the term data science, I'm not talking about uh computer science necessarily. I'm not talking about uh statistics necessarily. I'm not talking about um you know quantitative literacy necessarily, or mathematics. Um and I'm not even necessarily talking about uh what you might call, say, a domain expertise. Right. But what I'm talking about is the place where each of those concepts, uh computer science and programming, uh statistics um and uh some some kind of quantitative mathematics literacy, uh, and even um even the domain uh expertise, so knowing things about a particular context, it's the place where those four things overlap.

SPEAKER_01 4:38

Oh, I like that.

SPEAKER_00 4:39

Uh so I'm envisioning in my head, I know I've seen some, I've seen some Venn diagrams out there that have three, and I've kind of combined two different ones uh to make a uh a fourth uh circle in the Venn diagram. And so my my imagination has four circles with this Venn diagram, and it's the intersection of those four circles: computer science, domain literacy, quantitative uh literacy, uh, and statistics. It's kind of the the place where those four things meet up.

SPEAKER_01 5:11

Oh, I love that. And I love that visual of four, you know, four circled Venn diagram and the combination of all of those things. So what I'm hearing you say is it is interdisciplinary, right? So it's not just about um a single straightforward concept of, you know, here's what it is, and here's, you know, the indisputable content that everyone would label as data science. Um and the other thing that I'm also hearing is how it is applicable, how it is broadly applicable. There's a lot of complexity to what's happening in that definition of data science. So it makes us think about problem solving, it makes us think about interpretation, it makes us think about, you know, using mathematics, using computer science, using domain expertise and using statistical knowledge to actually, you know, do something with information uh that you wouldn't be able to do with one of those things alone. Right. So I I really like that idea of the complexity and the intersection all of all of those ideas. So so that being said, could you come up with like a a straightforward example, maybe from um, you know, middle school, or if you wanted to give a couple of different examples across across the K-12 span of um of applications, of what would be an example of something that that we would call data science that a a student in in K-12 education might engage with? What would that look like?

SPEAKER_00 6:50

So let me give this uh let me give this response. So I um one of the things that I talked about this summer, one of the things that I got to experience this summer was um the looking at different data sets. And that's really, you know, when you think about data science and the maybe the introductory part of this um for most of the the courses that I've read about or that I've seen described, the the concept is we've got this, we've got this data. We've either collected it or we've been given it, we've we've curated it from some other place. And that data is in a relatively rough state. But we've got some variables. Uh uh maybe we're looking at, and and one that I looked at this summer was uh some roller coaster data, kind of an engaging topic for high school students, right? We love roller coasters, we're adventure seekers, right? So thinking about what's the best roller coaster in the world. Sure. How do you even rank roller coasters, right? I mean, you've all we've seen the travel channel thing that says, like, you know, ranking the top 10, whatever. So how do you go about doing that? What is the process I might go through to do that? Um, and so it's the process of cleaning up my data. Maybe uh maybe the height, maximum height of a roller coaster is a variable that is of interest to me that I want to make sure that I include in my ranking of the high of the best roller coasters. Maybe the maximum speed uh is given and maybe that is something that I engage with in the number, you know, in the in the ranking of my roller coasters. Maybe it's the number of uh loops that it has, the number of times you go upside down, or maybe it's the number of Gs that you take, you know, the the maximum G force that's in there, or that I mean, any number of possible uh variables that I could include in this ranking of my roller coasters. And data science is the process of asking the questions, first of all, what am I interested in? Right. Uh, how do I come up with that data? Um, once I've gathered that data, can I get it in a way that uh that I can analyze it? So for example, one of the things we had to do was clean up uh the the data itself, maybe the entries for maximum height, uh, some of them include uh the units and some of them don't.

SPEAKER_01 9:10

Right.

SPEAKER_00 9:10

Uh so they they are treated as different types of variables by a computer program. Uh so there's, you know, I have to go in and make sure that everything is is all the same, and I can use some coding procedures to go through and and clean up my data. Maybe I've got missing data values, and so I have to kind of think about how I might manage that. So this is a larger project, right? Data science doesn't typically um you don't just kind of come in on Monday and you've got this is the problem you're working on, and then on Tuesday we're working on a different problem. There's uh you know, extended periods of time with a particular topic. And so that's an example. Another one might be looking at the Titanic data and making some conjectures about um the class ranking. If I know somebody happened to be a male and they were uh, you know, had a third class ticket, what's the probability of survival? Right. That kind of thing is also another exploration that you might be able to do.

SPEAKER_01 10:08

Yeah, I I want to circle back to that um Titanic data separate from our podcast recording because you've shared some hints about that with me before. I'm super interested in that. But but really what I heard you just describe is is is kind of like a cycle. And it kind of reminds me of like the modeling cycle or the problem solving cycle where where each each thing you described or alluded to there starts with a question like what's something that I'm interested in knowing? What's the best roller coaster in the world? Or um, you know, what what's the likelihood of surviving the Titanic based on, you know, some factor, some demographic factor? And although the descriptions that you gave were, you know, complex, and I I certainly when I think about roller coasters, I think about physics and I think about, you know, a lot of really complex level thinking that would be appropriate for high school. But I think this idea of that cycle of like, what's a question I I want an answer to? And then what data could I collect to help me answer that question? And then how do I, you know, clean up or organize that data in a way that I can make sense of it? And then how do I interpret it and then make some inferences? You know, that could be done even down in elementary grades, right? Like, you know, maybe a maybe a third grader is gonna have a question about do um, you know, do students that have pets have more chores? You know, there there could be a lot of different ways for uh a student to access this idea around asking a question, gathering data, uh visualizing and interpreting the data, and then making some inferences about it that don't necessarily have to involve, you know, higher order uh mathematics in in terms of the computation and that sort of thing. So really thinking about this as you know data science in the world is gonna look different than data science in school. But students can have experiences in school that helps them sort of mimic that process and get a feel for what's involved for data science that someone might do as a career field, for example.

SPEAKER_00 12:24

I think this this course or or at least this topic, this idea that's being talked about in mathematics, and we're gonna get into why is this important here in just a second, but just this this idea of data science in the high school or middle school or even elementary is really very much tied to the concept of data literacy, right? And making students um and helping students kind of uh be good consumers of statistics or of uh data or of graphics that are shown uh to them. So being able to make sure that they have uh an ability to look at those things and think critically about them and have some level of experience with how did that thing even get created? What was the the process by which that was created? One other thing that um I want to make sure that we're that we're clear about is often uh in a data science course, uh students may not collect the data themselves. They may just be presented some data. And then the question becomes, what are you curious about with this data that you know about? Right. What are you curious about? One of the one of the courses that I um I took and we talked about was they actually used um some local um city uh finance records. So they actually looked into uh the public record and got salary data and were able to kind of look at salary data and ask questions about local, like people that they local official type people that uh they were able to go in and ask questions about how does this work and why does, you know, what are the ways that this set of finances should be you know set up. And here's the question I have uh about the way that this is set up. And and so it you have an opportunity to really go into a lot of different places. Um, one thing that I and I'll I'll get off of this as quickly as I can, but one thing that I do also want to make sure that I I bring up is that data science is definitely not computer science. Um one of the things when we say, hey, we you know, you're gonna do some coding in a class, um, people get really a little bit afraid that all of a sudden we're gonna be just sitting down in front of a computer typing out a whole bunch of lines of code. Um, when really, in fact, a lot of at least in the introductory idea of data science, and I'm not talking about the practice out in the field and once somebody it becomes a data scientist, that's a whole different ballgame. I'm talking about what am I doing in a high school introductory data statistics uh class or a data data science type class? Um the level of coding that they're doing in that class is is much, much smaller. It's you know, more line by line, a little bit smaller. We may define a function, we may uh, you know, use a loop of some sort, we may use conditionals of some sort, but they're gonna be very, you know, it's gonna be small and not extra sophisticated kinds uh of code.

SPEAKER_01 15:32

Awesome. Thank you for that, for that clarification. I think that's really that's really important. And I think, you know, even we've we've spent, you know, 12 minutes already talking about what is data science, and I think it's still a little nebulous, right? It's still a little hard to wrap our hands around. Right. Um, but I appreciate some of the descriptions that you gave, and I think that'll uh kind of anchor our listeners as we move forward. In preparation for our recording today, uh some of the some of the content that I was reading I found super fascinating. I found this one statistic that in a 2018 study that was done, the researchers determined that 90% of the world's data at that time, this was four years ago, 90% of the data in the world had been created in the last two years between 2016 and 2018. So, you know, I think we all can certainly realize the impact that technology has had on our lives, that um, you know, just just the shift of of how we live our lives every day and the things that we interact with and the the enormous amount of data that's being collected, um, you know, not in a like somebody's watching me and and stealing my personal information kind of way, but just in, you know, being able to take lots of information and do something and uh, you know, use it for interesting purposes. So even thinking about um I I read something about uh health science in the health sciences field, like the ability to gather big, what, you know, quote unquote big data through medical records has allowed some healthcare organizations to actually identify risk factors for um, you know, life-threatening diseases or life-threatening medical events years and years ahead, just because there are these massive amounts of data that are available and the ability to organize and interpret them and actually identify, you know, wow, you're actually at a risk factor for a heart attack, and we can now pick up on those risk factors five years before you would have even had that heart attack just by analyzing all that data. So, you know, there's sort of an obvious like, yeah, it's important to have data literacy. This is an important thing for us to teach in school because it's a part of our of our worlds. But um, I'm I'm curious from you know the perspective of the courses you took this summer and your background research before today's episode, like why why should we be paying attention to this in the math in the K-12 math space?

SPEAKER_00 18:08

Well, I think you brought up um perhaps what I think is maybe the most important thing is and that's just the the sheer amount of data that our lives are driven by. Um and just you're right. The the there's been multiple studies out there that have shown the um the doubling rate of the total amount of data out uh in the world is somewhere between two and five years. Um so every between every between every 24 to 60 months, we double the amount of data that exists in the world, which is just um a phenomenal concept, um, that it continues to double at that rate. Um, and so that's that's maybe one huge reason why data literacy and really this concept of data science, because I know data literacy, there's there's a piece to that um where we're really talking maybe specifically about graphical representations and maybe sort of a civil uh interaction with data, right? In my my civic duties and really just thinking about how I read the newspaper and I look uh critically at studies and things. I think there's I think there's a piece of it that is just being data literate, understanding the terms and being able to hear that. But then I think there's also just a a piece of understanding the process of how we get to um the kinds of results that we get to um when we look at at data and studies and being able to uh pursue uh potentially a career in this. Um, I think there are many, many, many students out there who would jump at the opportunity to kind of analyze and think about, oh, critical, wow, I did have no idea that this pattern was out there in that medical, uh, in that medical journal and and or in the in all of this data that you that you brought up to me. And oh my gosh, we could predict this the heart attacks, or we can predict, or we can say, hey, this person's at risk for this or that thing. I think that's such a great interest for a lot of students, right? Who may not have typically called themselves uh to a STEM career. And this is one thing that I think we I want to bring up a little bit is this idea of man, doing data uh literacy and doing data science. I that's I I'm stretching it a little bit, I think, to say I'm doing data science, but just those concepts are going to be involved in so many careers that I wouldn't have predicted before. Right. Um, I wouldn't have necessarily assumed that I'm going to be needing the data skills in each of the, you know, in so many different careers that you really are going to need them in now.

SPEAKER_01 21:10

I totally agree with you. And I think it's it's interesting to me that this conversation around data science, I think we've, you know, made a compelling case for why this has entered the education space and why K-12 educators are talking about this becoming a an important part of mathematics teaching and learning for students, just to prepare them for the world in which we're all living, is certainly a compelling argument. However, you know, you and I have uh enough history in the world of education and classrooms and work with teachers, and the structures and policies around standards and assessment and accountability that it isn't just like, hey, we should do more data science and data literacy ideas in school so that students are prepared for you know the world that we live in. It's not as simple as that. If if if this is gonna become part of something that we do, then what's gonna give, right? How are we gonna, how would we fit it in? So there's a lot of conversation out, particularly I think in the research space and in policy, around whether this is important enough to make room for it, particularly in high school. And you know, we we we both have talked in our preparation for this recording about some of the critiques against no data science isn't appropriate, and whether it's um, you know, the the rigor, the actual rigor of what a data scientist would do uh or what a you know a big organization would do to interpret data like we've just been describing with these anecdotes is not appropriate for high school. And then also the argument of, well, if this is uh presented or offered as an alternate to calculus, then we're limiting the number of students who have the opportunity to be prepared for STEM fields. So I just want to call out that not everybody thinks this is, you know, an urgent thing. Not everybody thinks this is something we should be making room for in the already packed curriculum. Um, what are your thoughts about some of the critiques that you've read or heard about uh around data science? And how does that kind of fit into the bigger picture for you?

SPEAKER_00 23:21

So I think that's a first of all, the critiques are very interesting. Um, and they're thought-provoking because you know, they wouldn't be critiques and people wouldn't be writing them if they hadn't really thought them through, I don't think. Um and so there are some compelling arguments for hey, we need to be really careful. And one of them I was, you know, we were talking about before we even started recording was this concept of when we decide that we are going to give students credit, and I I'm gonna focus really on the high school um space here for just a minute, but when we decide we're gonna give students uh credit for um an algebra two equivalent, um, and there are there are states where that is now the case, where I can take a data science course as an alternative to algebra two. Um we have to be really particular and specific about what that uh credit really then means. Is it that the because data science is not algebra two. It's it's not. And and that is just that is just the truth. So what do we mean when a student has gotten an algebra two equivalent credit in high school by taking a data science course? Do we mean that that student has taken now a rigorous math course and this student is prepared to go on and do some more rigorous kinds of things in mathematics? They may need to go back and take uh an algebra two equivalent in college before they could pursue anything in the calculus strain of things. And let's be really very real if the student has chosen to go and do a data science uh career, a career in data science, they are going to need some skills in calculus that they will not have gotten yet. Um, and so they are going to have to go back and get those things, and that mathematics is going to have to be found founded for them. What we are saying, I think, what we can be saying, though, is that the student has been exposed to some level of rigorous material. That the student has been exposed to some level of critical thinking that maybe they wouldn't have gotten if they were in an algebra two course. Maybe they didn't get that that same kind of rigorous critical thinking uh over there. Um and so they are prepared for a college space. I I will go out on the limb and say that a student who's taken that uh course is is prepared for that just based on what they've um seen content-wise. But what I think we really just have to be very careful about how we define the course, how we say what a graduation requirement is that is met, um, and and what does that really mean and communicate to those that are reviewing my transcript.

SPEAKER_01 26:16

Yeah. And I think you've pointed out what I think is one of the most challenging things about bringing something new into the curriculum, bringing new conversation into the space is that there's it, it takes time to build this common understanding of what it is. And as I was listening to you talk about, you know, the validity of a data science high school course being a legitimate replacement for algebra two. Um I think the caveat there is, well, what data science course are you talking about? Like I think there is the very real potential. And again, the um the very real concern brought up by some of the opponents of this in the literature that we read in preparation is that this will become a watered-down, you know, alternate less than opportunity where, you know, flawed and biased systems in our educational um systems are set up to actually track certain populations, predictable populations of students into these courses, and therefore limiting their opportunities after graduation. So I 100% agree with what you said that um, you know, the level of thinking, the preparation for college-level work. Um, you know, this actually is, I agree with you, a rigorous, solid data science course. I think is a great experience for students who are going to take calculus or have taken calculus. Like I don't know that there's even necessarily one has to be a prerequisite for the other. I think they could go in any order. I think all of those things are true with the caveat that yes, if the data science course is rigorous and is aligned with, you know, these, these high-level ideas and not just um a bunch of, hey, let's look at graphs and say what they make us think and feel. Right. There is that very real potential for that to get to get watered down. So um certainly we're not promoting uh that the watering down, and we're not promoting uh using data science as a way to um break off access for certain populations of students. Instead, we see this as an opportunity to actually generate mathematical excitement and interest, maybe among students who haven't felt that excitement based on a more traditional um course sequence or a traditional mathematical exposure. So I appreciate you sharing that. And I think we'll we'll put some links into some of the critique articles uh in our show notes too, so folks can kind of explore those ideas a little bit more. I think it's important to look at both sides of the conversation to make some decisions about what is and isn't appropriate.

SPEAKER_00 29:08

For sure. And I was just gonna make one more comment before we move to our last little segment here, and that is that, you know, one of the courses I took this summer, they view uh this data science course as it's an entry uh course. So the students take it as freshmen or sophomore. Uh, but they they have made an explicit uh declaration that this is a uh a springboard course or a gateway course, where this course is the one that the students come in, they take this data science course, and then from there they springboard off and they can go and take uh a coding course, or they can go and take a statistics course, or they can go and take a little bit more mathematics uh to try to explore and get going in the math world. And this data science course, the things that they explore there are intentionally designed to give students opportunities and exposure to things beyond that course, but they're not designed to develop mastery of any of one of those things, but rather curiosity around each one of those things.

SPEAKER_01 30:16

Right. That's a really important, uh, a really important component of that. So thank you for sharing that. So, all right. So, with the time we have left, and again, we'll link to resources in the show notes, but for our listeners who are interested in learning more about data science and bringing some data science ideas into their math class, um, you know, what would suggestions be? Whether you want to share a resource or a way of thinking, you know, what how would somebody get started if they want to do more with this?

SPEAKER_00 30:42

Wow, there's so many different uh opportunities and things out there. I would highly encourage, especially folks who are thinking in the um the secondary mathematics courses. I I would go check out some of the universities who are sharing their data science, their intra-level data science courses with high schools. So I know the UC UCLA is doing that, especially uh in the state of Ohio. They've kind of really been driving the pilot there in the state of Ohio. I know that um the UC Berkeley has their Data Eight course that they have put out there for folks to go and to look at. Um and each of these use different coding languages. So it isn't necessarily that you've got to uh know just exactly what you want to do tomorrow when you go in. I think there's a ton of research to be done uh and go check out. Of course, uh Joe Bowler and YouTube have done a ton of work around this this concept of data science. And they've got a whole set of materials really developed, um, K-12, different ideas and things that you can do uh all there. So go check out all of their resources. And we've got links to the bulk of these uh in our uh show notes. But the main thing that I would say is that don't go it alone. Ask an awful lot of questions, try to find others who are doing the same things uh and and curious uh about trying to start these data science courses or what is it, what does it look like? Hey, we're gonna try to do this in our school. Don't go it alone. Um, I think there's plenty of folks out there wanting to do this and trying to find materials around doing this, or just even curiosity about doing this. Um, so definitely ask questions and find folks. And we'll post post some more uh resources uh in our show notes for that.

SPEAKER_01 32:34

Yeah, definitely. And I'm just gonna name a couple of uh resources that I have some uh personal familiarity with too. And one is um one of the uh university Cal State also has uh a statistics course and they offer, or not a statistics, a data science course that they offer free for teachers. It's course kata. Um and I saw a presentation on that that I was really impressed and excited by. So I want to plug that one a little bit. Um, the UCUBED resource has uh an interesting idea, particularly for uh educators who work with the younger students around maybe just think about, you know, number talks are a common instructional routine that uh, you know, elementary and middle school teachers use. So think about doing a data talk. And uh there's there's a a resource that helps explain what that might look like and how you might engage in that with kids. And one of the other resources I found that I thought was super interesting and accessible for, you know, sort of a baby step introduction to bringing these ideas into uh into the classroom. The New York Times has a series now called What's Going On With This Graph. And they publish some really interesting data in graphical representations that uh can spark conversation that can, you know, as you said before, students aren't necessarily collecting the data, but they're looking at data that's been collected and organized and asking questions and evaluating and um, you know, coming up with some interpretations. So those would be some of the uh, you know, the hot, easily accessible resources that I would add. And again, we'll have uh a long list of resources uh and information and ideas for our listeners to explore on this really exciting topic. So thanks for great conversation, Kurt. I think there's uh there's some real interesting potential here.

SPEAKER_00 34:21

Yeah, we may end up coming back to this one again here in uh in a little while because I think there's a ton to just be uh checked out. So um great. Thank you very much for today, and um we'll talk again soon.

SPEAKER_01 34:34

Well, that's it for this time. Be sure to check the show notes for the resources we mentioned and others you might want to explore. We would love to hear your feedback and your suggestions for future topics. And if you're enjoying learning with us, consider leaving a review to help others find us and share the podcast with a fellow math educator. See you next time.