Statistics can be a hard slog for the students taking the first, and sometimes only, jaunt through this quantitative minefield. But an understanding, or better yet a mastery, of statistics is a requirement for any successful scientific or research career.
Recognizing this tension, Purdue University professor of psychological science Gregory Francis set out to write an introductory statistics textbook that served both the needs students, whether stragglers or savants, and the professors trying to teach a roomful of both.
Francis already knew that most students don’t bother to crack their textbooks –assuming they even bought one. He also knew a fair bit professionally about human behavior, and how people interact with computers. So he took his textbook to the web, and the finished product, known as IntroStats Online (SAGE), debuted last year. “One thing that drove me to write this was the students told me they didn’t read their book, or it was obvious that they didn’t read the book,” he explained; IntroStats requires engagement from the student and lets the instructor know just who’s been reading and who hasn’t.
Like 25 percent of students actually read the book—and these were for introductory biology classes or introductory psychology classes—and it just struck me that this was so strange. We have authors who are taking huge amounts of time, to write textbooks that are really, really well crafted, and students in some cases are paying large amounts of money for these books, and they’re not being used
“I don’t know if that’s a carrot or a stick,” he said with a laugh. But he’s been offered a carrot of his own: the Software & Information Industry Association has nominated the textbook for a CODiE Award as the Best Social Sciences Instructional Solution.
IntroStats is Francis’ first “official” textbook, although he earlier had set up an online lab experience — the well-known CogLab Online Laboratory — where students did classical cognitive psychology problems to get a sense of what authentic experiments are like. “Students would read about them in textbooks,” Francis noted, “but it wasn’t quite the same as sitting down and experiencing it.”
Social Science Space caught up with Francis at the end of a busy day at the Human Brain Project at Switzerland’s École Polytechnique Fédérale de Lausanne, where’s he’s been on sabbatical since August. The American is one of the few psychologists taking part in a well-funded international effort to craft a computational neural network model of the human brain.
One of the most striking things about this textbook is that there’s no ‘book’ here.
It is an online textbook, and I think there’s a lot of advantages to having it be online. You get color for free, you get interactivity, you have immediate grading of questions, in some sense you can make certain that students are interacting with the material.
For years I used a standard textbook, and students would come a see me during office hours because there were a number of problems they had to solve, and they would say, ‘I have no idea how to compute this thing.’ And I’d say, ‘Well, if you just turn back three pages, there’s the formula right there.’ And they’d say, ‘Yeah, well I haven’t read the chapter yet.’ [Laughing] You’re in a class where there are homework problems related to the chapter — and you haven’t read it! This doesn’t make sense to me. What occurred to me with an online textbook, I can make sure that they’ve read it (or at least encourage them to read it).
So in IntroStats one of the features that I like—which I’ve never seen in any other online textbook—is that students log in, they go to a section, they read it, and the system keeps track of how long they’re on that page. And in order to get credit for it, they have to be on for a long-enough time (which isn’t that long, to be honest). At the end, they have to answer a couple of questions, very simple questions, and they get credit for the assignment. And so now if I ask that some of my students read sections 5.6 and 5.7 by Thursday, I know that they have.
In some sense it’s easy points for a student to get — which in a stats class is always good! Then I can be confident when I go into lecture that students have read those required sections.
I still have some old textbooks that I saved so I could look up things later. But with this …
I still have some old textbooks, too, but I never look at them. The Internet has changed the dissemination of information in so many ways. In particular for a subject like statistics, where someone can remember ’How exactly do I do a test for two correlations? — I can’t remember the exact details.’ Fifteen, 20 years ago, I would have looked through a textbook. Now I just go to Google and browse around a few sites I like.
I think information’s available to anyone who wants to get refreshed, particularly for this kind of information. But getting refreshed is not the same thing as learning the material in the first place. So I think IntroStats Online is good for the learning process. If people need to be refreshed the information is there, fine. They could find essentially the same information many other places, but it just doesn’t help them learn in the first place.
If you could go back to the beginning of the IntroStats process right now, what would you add?
Statistics is in a lot of flux right now. In psychology—and the textbook reflects this—it’s mostly about hypothesis testing, which is a particular way of making decisions to say whether something is different from what might happen just by chance. There’s a lot of concern about that particular approach when you build theories based on those sorts of analyses.
I think when I go back and do a revision I will attempt to address some of those sorts of issues in a meaningful way and reflect some of those changes that are happening. It’s actually quite curious, because in statistics most of the methods go back a hundred years or so, and so when you make a new version of a statistics books it’s not really about changing the content, it’s just about changing the presentation or the style, things like that. But that’s really not the case right now. I think over the next five or six years there’s much bigger changes going on.
That’s one of the things I like about the book being online, which is that I can go in and change it relatively easily. One doesn’t necessarily — although my editor might disagree!—even go through a formal revision process. It’s just HTML. You add another chapter, and there it is.
But you’re not going to merrily go about adding material …
In principle it’s trivial to do. I log into the website, I copy and paste the new text, I save, and done. The difficulty is, you don’t want to change an online textbook in the middle of a semester. When the instructor says go read section 5.6, and the students read it, then it’s different content from what the instructor thought — one has to be careful. From a technological sense the actual work is trivial, but from a reasonableness point of view for the instructor, for students, but also for the publisher and the author, it’s a little bit harder to see.
For a textbook, it’s easy — you bought Version 1.0 or you bought Version 2.0. But for a website it’s not so obvious. It seems kind of silly to host Version 1.0 and Version 2.0 simultaneously.
What are some of the challenges of this online textbook system? You’re maintaining a living entity, not looking over static books.
Here’s something I kind of knew from the previous experience with the labs. The most common issue people have is they forget their ID. So when you login you need your ID, you need your password. They forget these things — students forget them so frequently. Well, it’s hard to know exactly what percentage it is, it seems like a lot—of course I don’t see students who pop in and say, ‘I remembered my ID!’ — but there’s enough students who forget. And instructors do as well; instructors are no better than students in this regard. So you have to have a tech support system to enable that, particularly as people learn to use the new tech system and they’re confused by many things. I handle the tech support for the site, mainly because I need to know if things aren’t working right and they need to be changed quickly.
It is a very different style of textbook where you’re done, you sit back, let the royalties flow in, and maybe think about a revision in a couple of years. On the other hand, on the publishing royalties side, there’s no used textbooks—which are a huge problem for publishers and for authors. I have a colleague who wrote a very successful introductory psychology textbook, and he says with every new edition you see a huge royalty check, and then the bottom drops out pretty much within a semester because the used sales kick in. But for this, there are no used books.
There must be a lot of intro to stats textbooks out there. What did you think was missing, and what do you then bring to the table in IntroStats?
Really it’s just the technology, the online side of it, that I can monitor students’ activity in the book itself. I can say, read this section by this date in order to get credit for the class, and I can see if they do or they don’t. When I teach, that’s part of the course grade. I think particularly in a class like statistics, where a lot of students struggle because they’re not strong quantitatively, these are new topics, it’s strange, students need places where they get sort-of ‘gimme’ points, because they’re not going to do well on the exams and in some cases they’re going to struggle with some types of homework problems. I often look for places where I can say, ‘If you do the work, then you’ll get the grade.’
Technically, they don’t even have to read it. They can put the web browser up, play Grand Theft Auto in a different window, then once the time limit is up they can say they’ve finished the reading assignment, close the web browser, and go on their way. They don’t have to read it, but I think most students do. Anyone who goes to the trouble to set that up is going to go ahead and read it.
I looked into some previous studies about this, and apparently people have looked at how often students will buy the book, and if they then read the book, and it was almost never. Almost never! Like 25 percent of students actually read the book—and these were for introductory biology classes or introductory psychology classes—and it just struck me that this was so strange. We have authors who are taking huge amounts of time, to write textbooks that are really, really well crafted, and students in some cases are paying large amounts of money for these books, and they’re not being used. It just seems like a huge waste. And so I want to try and work around that.
It also helps them to stay up with things, instead of, ‘It’s two days before the exam and I better read chapters five, six and seven.’ Then it’s probably too late.