Wednesday, January 28, 2015

President of Coursera to give plenary lecture at AAAS meeting

The American Association for the Advancement of Science holds a meeting every February. This year the meeting is in San Jose, California. There are four plenary lectures [Plenary Lectures 2015]. Three of them will be given by prominent researchers who will be talking about science. The fourth is by Daphne Koller, President and co-founder of Coursera.

Coursera is a for-profit company offering "universal access to the world’s best education." What they mean by "best education" is MOOCs offered by professors at the "top" universities. There's no evidence to support the claim that the best undergraduate courses for the general audience are those given by professors at Stanford, MIT, Princeton, and Harvard. Indeed, there's quite a bit of evidence that this isn't true.

Daphne Koller is going to talk about The Online Revolution: Learning Without Limits. Keep in mind that at the end of every article published in Science (AAAS publication) there's a small notice stating that, "The authors declare no financial conflict of interest" or statements that clearly spell out the conflicts.

Why is AAAS asking someone to give a plenary lecture about selling online courses from someone with a clear financial interest in promoting her company?

Science education is important and there's plenty of evidence that universities are graduating students who don't understand science and aren't capable of critical thinking. There are hundreds of people whose main research interest is pedagogy and especially science education. They have proposed solutions to the problem and suggestions on how we should change the way we teach. Very few of them think that MOOCs are the answer and very few of them are trying to market their ideas for profit [ Reaching Students: What Research Says About Effective Instruction in Undergraduate Science and Engineering (2015)].

Why not ask some of those experts to address the AAAS meeting and possibly explain what's wrong with canned videotaped lectures? MOOCs are just ways of transferring the way we teach now to the mass market. But what if the traditional way we now teach (lectures) is wrong? Isn't that a question that the delegates at San Jose ought to think about?

Here's Daphne Koller giving a TED talk. Near the end she talks about the importance of "active learning" (student-centered learning). I'm a big fan of student-centered learning. She implies that having students take online courses from "top" educators is consistent with active learning but she's wrong. It's the exact opposite. In my opinion, it's a step backwards and by promoting MOOCs we are going to make it more difficult to convince professors to change the way they teach. Instead, the ones who are good at delivering traditional lectures will bring in money for themselves and their universities.1 They will be getting the kudos and the teaching awards instead of those who are paying attention to evidence-based methods and trying to improve undergraduate education.


Here's another video. It's an interview with Daphne Koller from June 2013. Listen to the first few minutes and you'll hear a different view of active learning. Here, she explains the concept of the "flipped classroom" where students watch online videos and then come to class to participate in "something that's much more engaging and stimulating, active learning" (4 mins). There's nothing wrong with that except that students could read a textbook instead of watching a video. The important part of the learning is what happens in the classroom and not what happens in the textbook or the taped lecture.

She also explains how they are going to make money (5-6 mins).

Daphne Koller is very fond of repeating the myth that the best courses are the ones taught at the top research universities. (She happens to be a teacher at one of these universities.) I bet she can't prove it unless she's talking about very specialized upper level courses.



1. For-profit companies like Coursera offer kick-backs to the professors and schools that contribute courses.

26 comments :

  1. "Why not ask some of those experts to address the AAAS meeting and possibly explain what's wrong with canned videotaped lectures? MOOCs are just ways of transferring the way we teach now to the mass market."

    Did you ever take an online course? I have done so multiple times (even with Coursera) and there are huge differences in the quality of courses. The good courses are not canned lectures and a quiz, their central features is a well attended forum section. There are some things that dooms ome MOOCs - the biggest mistake I see again and again is a "learn at your own pace" approach, where the entire course material is available from the get go. It seems like a good idea, but it means that the students will not be at roughly the same place in the course and discussions don't happen.

    I do not consider MOOCs as replacements for undergrad courses, but I do think they can be a valuable resource - especially for things that are not available locally. I regularly used R for quite some time, then spent some time on a course to get some more structured ideas about the language. That's not something available locally and it was a course that did have short lectures, but the majority of the time spent on it was spent coding. Generally quizzes gave you some dataset, explained what analysis you should do and then you had to write a script to get an answer. There was a forum section for each task, where you could post your script after the deadline for the assignment had passed. You'd then be comparing notes with the others.
    I do not think that an offline course would have been strictly better (you gain the option of discussing algorithms with others while you code, but lose the sheer strength in numbers while doing so).

    In short: The medium is not the message. A frontal style of teaching is not a necessary feature of a MOOC and the best MOOCs are precisely not traditional lectures as streaming video.

    ReplyDelete
    Replies
    1. Did you ever take an online course?

      Yes. I took one on evolution and it was horrible.

      I do not consider MOOCs as replacements for undergrad courses ...

      Good. Then I assume you disagree with those who think they will eventually make universities irrelevant.

      Delete
    2. May I ask which one? I've taken quite a few courses and as I said, the quality differences are huge.

      I agree that MOOCs will not make universities irrelevant. I do think that there is a decent amount of political will to make universities irrelevant, which is based on an aversion regarding basic research on the part of some political decision makers and on an attempt to divert funding from universities to other research institutions driven by academics in the latter institutions.

      This dual pressure is something I'm acutely aware of - funding for higher education is diverted from universities to tertiary schools with no research arm (or a research arm that is entirely focused on engineering) and funding for research is diverted from universities to research institutions that do not teach (again with a heavy emphasis on engineering, with some applied science thrown in). I'm employed by a university, but funding for my position comes from one of the latter (and thus I don't have teaching responsibilities).

      I do hold on to the Humboldian ideal of the unity of Bildung and Forschung. I think the modern university which aims to prepare students to conduct research and demands original research from them as part of their thesis is possibly the greatest invention in how we organize the creation of knowledge (i.e. Wissenschaft in the broad sense).

      But I don't see MOOCs as a big issue here. They can't do everything a university does, but they can do some things well and I think I have benefited from some of them. I'm currently in a MOOC on CUDA programming run by a computer scientist who started cooperating with nVidia when they developed CUDA and one of the people from nVidia. It's got somewhat tedious lectures, but the alternative is even more tedious technical manuals. And there are some programs that could be run on a GPU in a highly parallel fashion which are relevant to me. I don't care about a statement that I did the course. I do care about the possibility of implementing CUDA versions of some programs and turning that into a paper or two.

      I also do the occasional course purely for fun. Combinatorial game theory? I doubt I'll get a lot of use out of that one.

      Delete
  2. John Hawks is one of the professors who provides MOOCs. Don't know how his particular courses are, but perhaps it might be nice to try one.

    ReplyDelete
    Replies
    1. I know John Hawks and I'm sure that his online lectures are very entertaining, informative and packed full of all kinds of facts about anthropology.

      But that's not the point, is it?

      Delete
    2. But that's not the point, is it?

      Agreed, if the proposition is to try to replace the functions of traditional college and graduate school courses. I wonder to what extent that *is* the proposition, and to what extent MOOCs are wanting to appeal to people like me, long out of school, who have a general (vague?) desire to "learn more about this subject."

      Delete
    3. We could sell MOOCs as simply a cheap way of information transfer that doesn't require you to read a book. I have no problem with that. If you want to take a free MOOC on evolution and learn all the facts about evolution then go right ahead. Creationists can take such a course and feel confident that they understand everything there is to know about evolution.

      But if you want students to REALLY understand and accept evolution it's going to take a completely different kind of course. Decades of pedagogical research has shown that you don't change students' concepts by just dumping information on them. You need to engage them in thinking critically about their misconceptions and their deep understanding of the issues. You need active learning and it's no use pretending that you can do this remotely.

      Professors still insist on teaching science in the traditional way. They seem oblivious to the fact that we are graduating people with science degrees who are creationists, who won't vaccinate their children, who deny climate change, and who believe that homeopathy works. It's about time we woke up to the fact that we are doing something wrong.

      Delete
  3. Why is AAAS asking someone to give a plenary lecture about selling online courses from someone with a clear financial interest in promoting her company?

    You've co-written a textbook from which you get royalties. How is that in any way different from making money off of a MOOC? If someone asked you to give a talk about teaching biochemistry, would that be construed as unethical because it would be a venue for you to promote your book?

    ReplyDelete
    Replies
    1. ... would that be construed as unethical because it would be a venue for you to promote your book?

      What do YOU think? Imagine that AAAS asked me to give a plenary talk about the best biochemistry textbook. Would you praise them for being really scientific and clever?

      Delete
    2. No, but I would imagine that part of your talk might reasonably be about how a textbook can be organized to aid effective teaching and you might well point to your textbook as an example of a book organized according to your philosophy. That's what I expect Daphne Koller to do in her talk in regard to Coursera, not trash talk the competitors at edX.

      While you present her as "President of Coursera" you can't forget that she is also a Stanford professor with an extremely successful teaching and research career and not just some random Internet entrepreneur that AAAS invited for some obscure reason.

      Delete
    3. Do you really not see any difference between someone who is the president of a for-profit corporation with 50 employees and millions of dollars in capital investments and someone who earns a bit of royalty income from a book?

      she is also a Stanford professor with an extremely successful teaching ... career

      You never believed me whenever I claim to be a good teacher. Why do you believe her? Is there any evidence that she's an extremely successful teacher?

      ... and not just some random Internet entrepreneur that AAAS invited

      Why should AAAS be inviting entrepreneurs to give plenary talks in the first place? There are plenty of experts on science education who could have given such a talk. Why didn't they invite Bruce Alberts or Carl Wieman, or David Botstein, or Nancy Kober or Eric Mazur? Is it because none of them would have promoted MOOCs as the solution to the problems we face and none of them are out to make money from their academic research into teaching?

      Delete
    4. Do you really not see any difference between someone who is the president of a for-profit corporation with 50 employees and millions of dollars in capital investments and someone who earns a bit of royalty income from a book?

      Considering that Coursera offers courses for free and only charges for completely optional certificates, if there's really a difference, it's not one that puts textbook authors on a higher ethical plane, if that's what you are suggesting. Although to be fair, Koller has co-written a textbook as well.

      You never believed me whenever I claim to be a good teacher. Why do you believe her? Is there any evidence that she's an extremely successful teacher?

      Well, 1) I've taken her MOOC on probabilistic graphical models, so I've experienced her teaching directly and was reasonably impressed, and 2) She received the Cox Medal, which is Stanford's annual award for undergraduate teaching.

      Why should AAAS be inviting entrepreneurs to give plenary talks in the first place? There are plenty of experts on science education who could have given such a talk. Why didn't they invite Bruce Alberts or Carl Wieman, or David Botstein, or Nancy Kober or Eric Mazur?

      Do any of these people have anything new and interesting to say about education? From the names I recognize, although being distinguished scientists, they haven't taught undergraduates for years.

      Delete
    5. ... it's not one that puts textbook authors on a higher ethical plane, if that's what you are suggesting.

      No, I not suggesting any such thing.

      I've taken her MOOC on probabilistic graphical models, so I've experienced her teaching directly ...

      Okay. That's evidence.

      She received the Cox Medal, which is Stanford's annual award for undergraduate teaching.

      That's also evidence, although teaching awards often don't mean what you think they mean. Anyway, you have successfully supplied evidence that's she's a good teacher.

      Do any of these people have anything new and interesting to say about education?

      Maybe not so new since they've been trying to change the system for many years without success. But what they have to say is surely interesting.

      ... although being distinguished scientists, they haven't taught undergraduates for years

      I'm not sure how many of them taught an undergraduate course in the past year or two but they all have a lot more experience than you and I.

      Delete
  4. I think it is a mistake to to think of MOOCs as a replacement for classroom teaching.

    Among other things, they are the only available resource for poor people and third world people and working people. They are almost certainly the only resource available for working people who are just academically curious.

    ReplyDelete
    Replies
    1. It may or may not be true that this market exists and will be adequately served by online presentations.

      That's not the point. The point is whether MOOCs are the best ways to teach and learn. If not, then they should be marketed as a cheap alternative for information transfer that incorporates all the bad practices that we developed over the past century but not as examples of the way education SHOULD work in the future.

      Then we can tell all those people who can't go to university that MOOCs are not the best that universities have to offer but it's the only thing that can be delivered cheaply. Let them know that a second class education is better than nothing but don't pretend that MOOCs are the best we can do.

      And please, please, don't pretend that undergraduate courses from Stanford professors are superior to those taught at UNC, Wisconsin, Vassar, William & Mary, or Harvey Mudd. Or the University of Toronto.

      Delete
    2. I would agree that the methods you are experimenting with are more effective than lectures. But I spent most of my school life in classes nominally conducted using your methods. Small classes, open discussion.

      My experience has been that this can be better or it can be worse than lectures, depending on the teacher. One of the best classes I took was done entirely by lecture. One of the worst was interactive. The difference was the teacher.

      Delete
    3. The difference was the teacher.

      What do you suggest? Do you tell someone who does a good job of traditional lecturing that they should keep doing it even though we know there are better ways? And what about the bad teachers who practice active learning? Should we tell them that it's better if the give powerpoint presentations?

      There will always be good teachers and bad teachers. We should be aiming to teach by the best methods no matter how good the teacher is.

      Delete
    4. If I knew what to suggest I would be writing books. I am fairly modest except for knowing a lot of things that don't work, or which don't work quite as well as we wan them to do. For the record, I have an MA in special education, which is heavily into testing methods for effectiveness.

      I have no doubt that personal interaction is necessary or desirable. All special education involves interaction. Disabled people (mentally disabled) are usually not autodidacts.

      I have never taught for a living, but I spent some years in a small software company writing code and doing telephone support for a highly specialized system. I fell into the support role because I was the only one having the patience to deal with customers.

      For example, I spoke with a bank president who couldn't follow instructions to type an "*". He said his computer could only type an "8".

      I'm sure teachers have war stories about students who just don't get things that should be obvious. I don't know how to mechanize teaching. i wish I did.

      Delete
    5. Should read "which don't work quite as well as we want them to."

      Delete
    6. What do you suggest? Do you tell someone who does a good job of traditional lecturing that they should keep doing it even though we know there are better ways? And what about the bad teachers who practice active learning? Should we tell them that it's better if the give powerpoint presentations?

      There will always be good teachers and bad teachers. We should be aiming to teach by the best methods no matter how good the teacher is.


      I've had substantial experience as a student (kindergarten through graduate, then continuing professional education), virtually none as a teacher, unless you count training individuals to do particular jobs (which I wouldn't, at least not as relevant to the task of teaching groups of students about an entire subject area). But my very strong leaning, based on that experience, is to agree with Petrushka that it's the teacher rather than the particular method.

      I think it would be a mistake to take a teacher who does a good job using a particular teaching modality and try to train them to do something they may or may not do well in the name of "improvement." To me, this is like taking a golfer who is successful with a particular swing and trying to change it because everyone agrees a different swing should produce the best overall results.

      Now one might be able to take a bad golfer and make her better with a new swing. Or she might remain a bad golfer. We shouldn't expect miracles, whether in biochemistry, golf, or teaching. But why mess with success?

      Of course this then leads to the question of what success is in teaching and how to evaluate it.

      Delete
    7. Of course this then leads to the question of what success is in teaching and how to evaluate it.

      Exactly. But you seem to have the answer based on your own experience. Clearly you think that university teachers can do a "good job" using the traditional methods. You imply that those good teachers may not be able to improve by adopting a different methodology.

      I think you're missing the point. In your golf analogy we all know what the goal is—it's to get the lowest score in 18 holes. You imagine that what I'm trying to do is to change the golfer so that she has a better, more pleasing, technique. This might actually result in more boogies.

      That's not what I'm trying to do. I'm trying to convince the golfer to play tennis because all the really important stuff is on the tennis court and not on the golf greens.

      Being a good golfer is not going to count anymore because we're changing the game.

      Delete
    8. Hmm, dueling metaphors! Well, you have a much more detailed idea of what you are trying to do than I, so perhaps it really is tennis and you'll succeed.

      I'm not optimistic. No change in teaching methods I can remember ever resulted in a general improvement in teaching ability. There was the "new math," and "phonics," and that was just grade school.

      In fact, my experience of university and graduate school teachers was close to the opposite of what you are assuming. A few were quite poor, the majority were mediocre to fair, and a few were excellent. The distinction did not seem to fall along the lines of teaching methods. Yes, I'm saying we probably shouldn't mess with those doing a good job (again, the problem arises of defining what that is); I'm also saying I hope but doubt it will help the vast majority.

      If it did, it would be the one time in my experience that it's happened. Have you ever experienced this (a general across the board improvement in teaching because of a methodological change)? If not, do you suppose it's because no one's ever been smart enough to come up with good ideas about teaching before?

      Delete
    9. A few were quite poor, the majority were mediocre to fair, and a few were excellent. The distinction did not seem to fall along the lines of teaching methods.

      I don't expect it to. Do you?

      If it did, it would be the one time in my experience that it's happened.

      You do realize, don't you, that this is not an argument against what I'm proposing? If you agree with the notion that change is needed then whether you ever experienced it in the past is irrelevant. If you disagree with the strategy I've outlined then it's also irrelevant. There's only one situation where your argument makes any sense and that's when you agree that change is necessary but you don't think it's possible because you've never experienced it.

      Nobody said it was going to be easy but lots and lots of university teachers have already changed. I have.

      Have you ever experienced this (a general across the board improvement in teaching because of a methodological change)?

      My granddaughter is in a nursery school that borrows heavily from Montessori. It is far, far, superior to the nursery school I attended and better than the ones my children went to.

      I went to a school where we sat in rows with assigned seats. Today's open classroom seems to be superior in many ways.

      Nothing has changed in the universities since I took classes in the mid-1960s but the public schools and the high schools are completely different from what they were 50 years ago. Almost all those changes are for the better.

      Delete
  5. Larry, I don't know why you're so hostile to MOOCs. Like face-to-face courses, there are outstanding MOOCs and just-OK ones and bad ones. The good ones typically do manage to incorporate active learning, despite the on-line format.

    ReplyDelete
    Replies
    1. I know you've spent a lot of time developing your online course in genetics and I'm sure you think it's as good as the lectures you give in the classroom.

      However, over the past decade my friends and colleagues who specialize in pedagogical research have convinced me that there is a much better way to teach. It requires continuous give and take in the classroom as you guide students to an understanding of fundamental concepts and encourage them to develop critical thinking skills.

      The students have to interact with each other and take the lead in learning. That's why it's called "active learning" and "student-centered" learning. The instructor is a guide and facilitator and not just a conduit of information. In this form of education the instructor is an active participant in the classroom discussions. Instructors and students are part of a team with a common goal ... the goal of understanding.

      You never know for sure where the class is headed. You have to "go with the flow" when the students reveal problems in grasping difficult concepts or sorting out ideas. You have to take advantage of misconceptions when they arise by getting students to deal with them while arguing with each other. (The instructor guides by asking provocative questions or mentioning inconvenient facts.)

      These objectives cannot be accomplished if the instructor is viewed simply as the "sage on the stage" delivering a powerpoint presentation to the ignorant masses. That places the instructor in a much different role with respect to the students. The see her as the authority figure in the class and not as a partner in their journey.

      MOOCs reinforce that traditional view of the instructor. Without the personal face to face contact that's essential in the modern view, there's no way to succeed. It would be like asking Socrates to teach via Skype.

      It's hard enough to convince my colleagues to adopt evidence-based teaching but the task is made much harder by those who claim that they can all be replaced by online courses. I think MOOCs are a step backwards. They instututionalize the traditional lecture-based formats that I want to replace.

      Delete
  6. If only Feynman had been taught in a flipped classroom...

    ReplyDelete