TSDI - Unit 1 - Expert Panel

Video Transcript

Hi, I'm Dr. Hollylynne Lee, and I'm here today with some of my good colleagues and friends who are clearly experts in Statistics Education. So thank you so much for coming and joining us here, it's fun to have you all together here around the same table. They are serving as our expert panel for our MOOC on Teaching Statistics Through Data Investigations and I've known each of you for years, and you've each done many contributions in Statistics Education so I'll like to give you a moment to introduce yourself and to tell us a little bit about an overview of what you've been doing in Statistics Ed. Chris I'm going to start with you.

C: Okay, well I'm Christine Franklin, I'm on faculty in the Statistics Department at the University of Georgia, and I would say that probably my last 20-25 years I've devoted my career to trying to infuse more statistics in K-12. I would say just to give you a brief summary of some things I'm very proud of in terms of my work, one would have to be the pre K-12 GAISE framework that was initiated by the American Statistical Association. Also, I've been very active in AP Statistics serving as a former Chief Leader. So I would say that my real passion is K-12 and also teacher preparation at K-12.

H: Okay, thank you. Susan, let's go over to you.

S: My name is Susan Friel, and I'm at the University of North Carolina School of Education on faculty there. It's hard to figure out where to start, but I was an elementary teacher at one point in my life many years ago, and then moved over to Leslie College in Cambridge, Massachusetts and got involved, I am a curriculum developer. So I got involved with developing curriculum, and the curriculum I got involved in developing was called "Use Numbers" which was funded through the National Science Foundation as one of the first statistics curriculums that was actually designed for use with elementary students. That later became part of an elementary math project that was funded through NSF, and I went on to continue doing work in statistics in a lot of different areas and working with technological tools, they intrigued me in terms of learning about that. I moved to North Carolina eventually to head up the Math Science Network originally in North Carolina, and as part of that we had a multi-site NSF grant called The Teach Stat Project, and I worked with two statisticians at Appalachian State and we developed a three week professional development program for teachers across the state. They were trained as leaders and then we developed a one week leader training where they learned to teach other people, and I like to think that Teach Stat is still going on in terms of some sites and teaching teachers how to teach statistics.

H: I still draw some of your tasks.

S: Right. It's a really good project, and then along about that time I started working with people at Michigan State and University of Maryland to do a middle school curriculum called The Connected Mathematics Project, and I have been the lead author on the Stat Units in that curriculum for the last three iterations of it as well. So I would have to say the statistics has consumed me.

H: Good.

W: My name's Webster West, I'm a Professor of Statistics here at North Carolina State University. I've been active in Statistics Education for 20 years now; I did the calculation on my way over today. I started my efforts in Stat Ed with the advent of the internet in the 1990's, I was sort of a programmer at heart, and I began developing some internet resources and technology resources for illustrating basic statistical concepts. And what started as sort of a hobby grew over time, and using some partial support from the National Science Foundation I developed a stat package called Stat Crunch that is now used very heavily in Statistical Education, particularly, at the college level about ½ million students a year are using it now. In addition to that work on the technology side, I've also done some curriculum work, once again, with the support from the National Science Foundation I've been working on developing randomization and simulation methods and building some materials to use that--incorporate those types of techniques into the introductory statistics course. So that's some of the things I've done.

H: So now you have met our experts and we're going to have a series of questions for them and we're going to start with 'What is statistics?' Webster, we're gonna bounce that to you, what is statistics?

W: Okay, I'll give you the same definition I give my students, I say, "It's the science of data." Not to be confused with data science, but, well maybe it should be confused with data science, but yeah I see it as all things data. So from collecting data, summarizing data, and if appropriate, making decisions based on data. So really anything that has to do with data, not necessarily just numbers, but text, there's lots of things that qualify as data. So, anything to do with data I say falls into the purvey of statistics.

H: All right. Chris, do you have anything to add to that?

C: Well, actually that's a great answer Webster. I'm going to put a plug in.

H: Okay.

C: For my textbook, that's an introductory statistics textbook, and the way we define statistics is we say "It's the art and science of learning from data". And I really like to put the word art in, as well. I think one of the important things about statistics and being statistically literate is knowing how to communicate, how to communicate your results, writing orally, and so I like to think that that's the art of statistics. It's also quantifying variability, and as Webster said, it's no longer just numbers and a static worksheet the way we used to think about statistics. Statistics now is sort of a very dynamic complex structure such as pictures and sounds. So I think it's a very exciting time to be a statistician and to think about how we're surrounded by data every day.

H: Yeah, yeah. Susan, do you have anything to add to that?

S: I just had a little bit, I think of it as making sense of data, what your definitions are excellent, but the one thing that's always fascinated me and I started my students thinking about is, what questions do you have? So, what's a typical height, what's...how come we can ask those questions, what is it in the world that actually makes things, we be able to answer those questions? And so, statistics is the work that helps us analyze that data, but isn't it amazing that data has patterns to it and we look for them, and when we look for them we can say things and sort of allow us to generalize about the world and what's in it, and I think that's very neat.

H: Right, right, and I think just extending on the idea of it being an art. I think technology really affords us that art side of it because we have all this....we have great ways of visualizing our data that go way beyond the standard things that we used to be able to do with paper and pencil. As new ways of visualizing data, different software packages are coming up with these facilities to let us do this, and I think that's exciting.

So, thinking about statistics, it is in our curriculum, and particularly, all across the world, most countries have documents that require some teaching of statistics in the K-12 schooling level. So, here in the U.S. we have the Common Core, and...but why should we be teaching statistics at that level? Susan or Chris?

C: I'll go first.

H: Okay.

C: Let me give a reference. One of my responsibilities at the University of Georgia is that I'm the undergraduate coordinator in my department, and when I first began this job 15 years ago, I may have had at most fifteen statistics majors, and as of this past week I'm advising over 120 statistics majors at the University of Georgia, and that doesn't include the 100 statistics minors that we have at the University of Georgia, and I think this just tells us the importance of being statistically literate and knowing how to use...and how to have good statistical reasoning skills. It can't start with an intro course at the college level. We need to start at kindergarten helping our students understand the data that surrounds them and how to reason with that data, and hopefully, by the time they graduate from high school they'll have the skills that we're trying to teach in our intro course now. We are surrounded by data, and I think one of the things I try to help my students understand is how much data they actually generate every day, and I think they're just totally amazed, and to think about who's using the data that they're generating every day, and how's that data being used? What's the ethics involved with the data that they're generating? Just simple things like this, picking up a newspaper article, knowing how to ask the right question about the statistical results that are being reported in the media. So I think it's just...we're doing a disservice to our students if we don't introduce them to these statistical reasoning skills at K-12.

H: Susan, what do you think?

S: Well, as I said I had helped write an elementary curriculum many, many years ago, but it really was the National Council of Teachers of Mathematics; in both their curriculum documents, the curriculum and evaluation standards, and then the principal of standards, I think, was the better one of them in terms of recognizing the place of statistics; and the American Statistical Association and the GAISE project accompanying that, and so they influenced it as well. So we literally had a K-12 strand of how students should be learning statistical ideas, and since my focus is middle school, I was always very excited because usually the first year our middle school curriculum was building off of the things the students had learned in the K-5 curriculum in terms of data. With the Common Core there has been a little bit of a confusion about what we should actually be doing, and even though many of us have attempted to influence it, data in the elementary grades has definitely lost ground. So students in turn they come into middle grades without a lot of data experiences, and what you might find interesting, I was just talking with some of my elementary teachers I work with and they said, "We really miss that probability and data have disappeared from the elementary curriculum" from their standard course of study. They said, "There were context in which kids could problem solve and apply number knowledge". And so, probability was good for fractions, data was all kinds of analysis with numbers sense where kids could really make numbers in. None of that happens any more, and so, it's at the middle school level and when you start 6th grade, I like that they've pushed us to think about variability, but they've pushed us to think about it without necessarily thinking about the base, but also you have to try to figure out the best problems to make sense. So the middle school's definitely more sophisticated content, which is probably appropriate given what we need to know about statistics, but how to help kids understand and make sense of it is the big issue.

H: Yeah, yeah, so with the college level Webster you've got lots of experience, particularly, at that introductory statistics course for, it's often freshman, and a lot of majors and programs actually require at least one statistics course. Why is it so important at the collegiate level for these students to be engaged in this?

W: Well, I would go back to what Chris said, data is everywhere, and one of the things I do in my class, and the very first day is I do a survey with the majors that are there, and ask them how they think statistics might be useful to them, and we really get involved in some discussions and most students these days are at that point already understand it, and this is sort of important, there's a reason I'm here, but you know, as they go through their academic career I always tell them "The odds are, that they're going to do something in college relative to their major that involves statistics." And once they get into the job world they're like, you're going to be even more surrounded by data, and having good quantitative skills, good statistical skills, it's important in almost every profession, so it's an easy sell these days.

H: I think so, I agree, I agree.

C: I might add to what Webster just said, I think one of the reasons that we're seeing such a big increase in the number of statistics majors and minors, and a lot of these majors are double majors; so they have another major and they're picking up statistics as a major, they're starting to realize it is the statistical skills that will make them marketable. And I work closely with our business school at the University of Georgia, and when the big business recruiters are coming on campus, more often than not, they're more interested in seeing my students than they necessarily are seeing the business students. Because they're wanting to recruit students that have those quantitative statistical analysis skills, as well as, a lot of the soft skills, the communication skills, the technology skills, the computing skills, which of course, all of our majors, hopefully, are developing through our curriculum as an undergraduate major.
H: So I want to go off of the idea with the technology, earlier I had mentioned about technology affording us these opportunities to really visualize data in interesting ways, but one of the things that technology gives us access to is data itself. With all this different technology and the internet we can access so much rich data. What's the role of using real rich data in our courses? What's the value in that? Anyone want to...

S: I'll start.

H: Yeah.

S: I was thinking last evening about some of the work that Paul Cobb and Kay McClain did in terms of data analysis at one point, and one of the things they highlighted that when kids actually worked with data they had if they...so for me there's two ways of getting data for elementary and middle, probably secondary and college is different; provided data sets that you provide that are about real world settings because that's the only kind of data I know, and typically, when I do provide the data sets they are real data; I've collected them somewhere. And created data sets where they collect their own data and people argue for younger students and middle school students create, you know, collecting their own data, but sometimes that's very time consuming and it's difficult for classroom teachers to manage it. If you use provided data, Paul and Kay have said, "There's a process you need to do with them to engage them into the data set before you even do any analysis", and they talk about having students look at the data, think about how could it possibly have been collected, what was actually going on when you did it, and so they have students engaged in the data set really understand it, make sense of it, if possible, think about what if they added their own data to the data set, what would it look like? And so, even before you do any kind of analysis they have to unpack the data. And typically, if you haven't collected your own data and gone through that process, you have to do the unpacking because you don't make sense if you're presented with a table of data, and you don't have any chance to sort it out, you don't make sense of it, and so, the statistics of you do become more procedural than it is sense making.

W: Okay, so I would say building on what Susan said, I provide about 40 data sets to my students in the Introductory Course here at NC State over the course of a semester. And when I think about real data, you know, I think of real data as typically having reasonable size in this day and age, so not your typical textbook size sometimes. So if my data sets are several megabytes in size, 68,000 rows, a few hundred columns in a couple of cases. So I think that's very important for students to get that feeling. I want them to have the feeling of being, and somewhat drowning in data sometimes because that's very real, and you do have to sort of unpack that information and think carefully; it can be overwhelming about what's actually in that data, what measurements were made and things of that nature. So I really do spend a lot of time in that regard in my own classroom. Technology, I think that was sort of the starting point here right, you know, to consider a few megabytes of data without technology is impossible. So there's got to be some way of dealing with that, technology is the tool that statisticians use to deal with data. And any course that doesn't use technology in my mind is not really giving students a very...a real image of what statistics is about.

H: I would agree with you.

W: So I think that's very important, and the other thing that I think technology does for us is it allows for us to collect data from our students so in my classroom students provide data live, and we look at the data that was generated; the nice part about that is its...they understand how the data was collected, they were a part of it, and students are always more interested in any data set they are a part of and they want to know what dot they are on the screen, and then my students subsequently, use that same technology then to collect data for their project later. So, once again, seeing that process of collecting, analyzing, and making some decisions, technology allows them to do that in the integral.

C: Maybe I can make a couple of comments, I'll follow up with what Web has said at the collegiate level. Let me plug one of courses that we offer for our undergraduate statistics majors at UGA, we have a year long capstone course, and we're very proud of this course; they learn cutting edge techniques, they actually work with a client on campus on a project, it's a consulting project, they have to write a report, they have to give a poster presentation, but I think one of the big skills that they learn from this course is how to work with messy data. Because invariably the data sets that these clients are bringing are not in a usable form, and those types of skills are so critical, and our students often talk about how, once I got a job in the real world that's the only way I deal with that, is trying to clean up the data sheets. So I think the technology is just, it's invaluable. The other comment I wanted to make was in respect to K-12. Now this is anecdotal evidence that I'm giving here, but I've had the good fortune of working for many years with teacher preparation professional development with our teachers in Georgia, and I can say, initially, when I would first start working with them, they didn't even want to be at that professional development because they didn't want to teach statistics, but Georgia had it as part of its curriculum at the elementary level, as well as, middle and high school well before Common Core. Two years later I would see these same teachers and the first thing they tell me is "The only time I'm really able to really engage my students in the math class is when we're teaching the statistics." And it's because the students...it's real world, but also they're beginning to see a connection between why they're learning the mathematical skills. H: Yeah, they're applying some of that.

C: They're applying some of that, and it's just, if we can get our teachers comfortable at K-12 with the idea that they can teach statistics, they eventually begin to see, as Susan said earlier, the elementary teachers, "Why did they take this away, this is when our students really gain number sense?" It's just a wonderful way to integrate the curriculum to make mathematics a real world to the students. H: Right, right, and I would just add to that as far as with the importance of real data. I think it's really important to gradually give students access to data that is bigger and bigger, more cases, as well as, more variables. Not everything is univariate and vivariate, and I have research, there's lots of research evidence that students can coordinate multi-variables; they can look at and think about the affects of different variables and it affords them these kind of question posing of, "Oh, I've seen this relationship, I wonder if this variable or attribute impacts what I'm seeing here?" And the more that we can get them engaged with data sets that are bigger and have more variables in them, the richer the kind of expiration that they're going to be able to do.

S: Can I add a footnote on that one too? And real data sets doesn't mean created data sets from somebody, it means real data that you've gone out and collected somewhere.

H: Right, right, yeah. I agree.
C: Which is not always easy to do.
S: No, but it doesn't mean you contrive the data set.
H: That's right, that's right. But in this course you're going to actually get access to lots of real data, and data that has been created and stored for you so that you can actually use it in your course.
Thank you.