Interest In On-Line Statistics Courses?

You might have heard what happened to when Stanford professor Andrew Ng put his machine learning (a practical kind of statistical modeling) on line. He expected mild interest. One hundred thousand students signed up.

This was enough to excite even the New York Times, which dispatched Tom Friedman to investigate. He called Ng’s success a “breakthrough”, which it certainly is, in its way.

Friedman also puts us on to a new company called Coursera, which offers courses by professors from well known universities, in much the same spirit as Ng offered his. “The universities produce and own the content, and [Coursera is] the platform that hosts and streams it.” Too, many universities already have on-line courses housed on campus.

Several things. Stanford pays Ng’s salary. This money is only partly for teaching, and more so for Ng to publish papers—which contain the material which will eventually be taught. If Stanford didn’t pay him, he would have little time to think of what new to say. Also, Stanford rightly owns the content to Ng’s lectures. Pay for work, etc. Same thing at Coursera, which hires out the professors for a cut of the pie. All well and good.

Some courses are ideal for the web. The closer the course content is to cookbook recipes, the apter. I mean no disrespect. A courses which shows you how to install and run a certain program is nothing but a series of recipes, a well marked path with milestones and a known destination. Basic machine learning fits this scheme. As do the courses offered at Coursera: Algorithms, Calculus, Introduction to Logic, Vaccines, and Securing Digital Democracy (electronic voting schemes). Recipe does not mean easy.

But maybe not statistics. Unless you want a course in classical frequentist thinking, which is cookbook all the way. Coursera has one just like this. Taught, like it is in many places, by a psychologist (who I’m sure is a nice guy).

Don’t see any courses in history, poetry, literature, high-level philosophy, and the like; all classes which are not amendable to multiple choice testing. These are courses which don’t necessarily have an end, or have different possible ends depending on the mixture of students, or which require students do a lot of writing and talking.

On-line, it’s just as easy for a professor to grade one multiple choice or only-one-correct-answer test as it is to grade 100,000 such tests. But only if the class is recipe-based. One professor could not read through 100,000 essays, or listen to as many presentations. In fact, for these kinds of “free form” courses, he may not be able to handle as many students on-line as he could in person, since the material delivered remotely introduces some level of ambiguity and slows down interaction.

Student interaction for recipe courses, which are (and should be) more popular, is greater, because if the recipe says that “at time T, add X cups,” then it’s likely many students will have this information at hand when they are queried at some forum. This is not as likely in the free form class, where the answers are rarely as firm.

The way I run my introductory class is free form. It is also not a regular course in statistics, but Bayesian from the get go, and in the predictive sense (“all parameters are a nuisance”) regular readers will understand. This marks it an oddity. I get away with it because of a certain rare confluence of events. But it would never fly at most universities, where professors at their professions are more conservative than Rick Santorum (“What! Teach Bayesian probability before frequentist? Never!”). Just ask any professor how easy it is to introduce a new course into the system.

I lecture, but not in contiguous blocks of time. I ask the students lots of questions and frequently. The answers they give provide direction for the course. Students come to the board and work out various matters with themselves, guided by me. I have students collect data (which fits a given paradigm) on any subject which interests them. This a wonderful way to maintain interest, but it limits the use of canned examples, the use of which would free up time.

I also have to spend a lot of time walking people through basic computer tasks, especially R. As a final exam, each student presents a talk on their subject as to an audience presumably unaware of statistics (many use data from their workplace, which they later show their bosses, reportedly to good effect). They must describe their interest, the data, the pertinent questions, show pictures, explain how they quantified their uncertainty, and finally detail how they would check all when new data arises.

Could this work on line? I’m skeptical, but intrigued. Places like Phoenix University push thousands of students through their pipes, and not all the classes are recipe-like, so maybe it can be done. All ideas welcomed.

The last difficulty is “credit.” Some courses earn credit as normal classes. But the free courses universities and Coursera offer come with nothing except a pat on the back or perhaps a letter stating that the student made it through. Of course, I love this trend away from formal “certification” and towards actual love of learning. Seems to work best for recipe courses, though, whose students actually want to bake a cake.

The “free” part doesn’t hurt when attracting students: a fine plan for behemoth institutions; wouldn’t work well for little guys like me.

Ignore below

Test of latex. Should be pretty: sin(x) \ne \int e^x dx, \Pr(x|e) = 0.5[\latex]


  1. I’m taking the Coursera ‘Compilers’ class, which is pretty good. I’m not sure how it’s implemented, but the videos can stop for a multiple-choice problem. That wouldn’t give the interactivity you want, but it’s a great opportunity to test understanding of a new concept.

  2. Dr. Briggs:

    The online courses I have taken typically have an instructor with a virtual whiteboard and audio feed, with a chat interface for student questions. Performance is measured with multiple choice testing. This method does not seem to lend itself to your preferred teaching mode.

    In recommending technology, I can only speak to my preferred learning mode. I have two engineering degrees, twenty years in the field, and very little scheduleable time.

    I prefer to RTFM, before bothering a domain expert. If I can do this disconnected, with my own intermittency, then at worst I only consume my own time.

    I prefer reading static media (including long form) over streamed media. If I must watch streamed media, brevity is mercy.

    I prefer graphics over symbolic math (e.g., Venn diagrams, areas under the curve, flowcharts). If I can explain the lesson to someone else using only a pencil and a napkin, then the lesson promulgates itself.

    Similarly, if the lesson uses a metaphor (your fashion-model for statistical-model lesson comes to mind), then I can convey the point orally in a staff meeting or in an elevator.

    Feel free to use low humor to make points. I’m partial to “Don’t drink p.” That will be the takeaway.

    My advice: Choose a brief point to be made. Use a graphic to make the point, perhaps with surrounding text. Otherwise make a one-minute video, using a whiteboard, and make just that particular point.

    Sell longer versions of these videos, where you make more complex points, or string together points in a bigger lesson. Reference the longer videos in the shorter videos and the static media.

    Finally, use all of these as promotions for the interactive teaching service you eventually select.


  3. I love the approach to teaching that you describe, but, for better or for worse, don’t believe it scales well.

    Jazz music works very well in an intimate venue because the musicians are “in contact” with the other musicians and the audience in real-time: they all feed off each other. Not having been in your class, I can only imagine that the dynamic is similar. Your students are, in a way, both audience and ensemble, and I bet you would lose a lot of what you do in, say, a 1,500 seat lecture hall.

    You might be able to use a similar concept: that of discussing and solving real problems, in an on-line format, but I would venture to say that you might find it difficult to produce the same result.

    This is just my naive view, though, and certainly wouldn’t wish to discourage you from trying!

  4. We are in very early times for on-line teaching/learning. So far, the emphasis has been on more students (or in modern terms — more impressions) per lecture which is like building larger auditoriums.

    From the 2010 Annual Letter from Bill Gates …

    The foundation has made a few grants to drive online learning, but we are just at the start of this work. So far technology has hardly changed formal education at all. But a lot of people, including me, think this is the next place where the Internet will surprise people in how it can improve things—especially in combination with face-to-face learning. With the escalating costs of education, an advance here would be very timely.

    … and from his 2011 letter …

    The foundation is funding the development of online tools to help both K-12 and college students learn. Pioneers like Sal Khan are already showing how effective online tools can be. His website continues to grow its library of 2,000 short instructional videos on topics from basic arithmetic to complicated subjects like biology and physics. The videos are a tremendous resource for students of any age.

    Sal’s vision for how technology can improve learning is broader than just videos. With support from the foundation, he’s been able to expand his site to include online exercises that diagnose weak spots, pointing you to additional material to fill the gaps in your knowledge. Also, Khan Academy is creating on online “dashboard” to help teachers use the site as part of their curriculum. The dashboard tells the teacher how each student is doing, pinpoints where they’re having trouble, and suggests explanations and exercises to help.

    Although it is clear that online learning works for strongly motivated students, we need to learn how to blend classroom learning and online learning, particularly for younger and less-prepared students. As these projects develop and we start to answer many of these questions, I believe technology will let us dramatically improve education despite the budget constraints.

  5. The Udacity course “CS373 – Programming a Robotic Car” uses the virtual whiteboard, interactive Python and starts off Bayesian. It’s free and you can start anytime. You can proceed at your own pace or just check out a few units to see how it works.

    I went through the first Unit of the Udacity “CS101 – Building a Search Engine” course to evaluate it as a first programming course for a young nephew. I did not like anything about that course. From my prospective as a Software Engineer, the sequence of material and the presentation style were just not appropriate for a beginner. I had high hope after the CS373 course.

    I recommend that everyone at least try out one of these courses to get a feel for the current technology and techniques. No cost, no commitment necessary.

  6. First, as I’ve already mentioned, I would be more than willing to pay you cash money for some lessons. I suspect there are others like me. I think you would find it rewarding too, as your students would be people who are interested exclusively in learning, and not neccesarily in a career path.

    Ive used Khan academy quite a bit, and Apple offers a lot of different courses through iBooks. Wikipedia is great for something’s too. Wolfram Mathworld is a good resource, and NIST has an online handbook of statistical practices for engineers which is very well written and extremely easy to use. The past 5 years have started a golden age for the autodidact.

    There is a guy on YouTube, mathematicalmonk, who has some fantastic videos. I’m not sure who is he or where he comes from, but if he ever reads this: “Thank you!!!”

    I think you would be very popular if you offered up some online courses. Count me in already.

  7. During my undergrad days I am quite sure that many lecturers had been replaced with robots; the lab staff were generally human so an education was possible– not easy, but possible.

  8. Let me put on my cognitive psychology cap for a few moments. First, multiple-choice questions can be as subtle as a sonnet, or as wooden as doggerel. That is, a fair amount of research indicates that the multiple-choice form itself in no way precludes discriminating questions about one’s knowledge of “history, poetry, literature, high-level philosophy, and the like.”

    For example, I daresay the eons-ago score I made on the AP Music Theory test (I specifically remember having to analyze a Hindemith vocal motet) was reproducibly higher than somebody’s who didn’t know much about music. How much more may we ask from any test than it more or less accurately tell the difference between the knowledge of test takers?

    Also, it is not accurate to say that essays cannot be machine scored. Even in the 1990s Thomas Landauer and colleagues were using Latent Semantic Analysis and selling software that they could prove would score free-form student answers and essays as reproducibly as human graders — as long as the domain of knowledge were tightly specified (such as giving the machine the contents of Chapter 7 of the student’s Biology text, with teacher questions drawn from that content).

    The key to it all, then and still now to some large degree, is comparing the ‘closeness’ of one ‘document’ with another — the same thing that Google does today.

    And of course IBM’s ‘Watson’, answering Jeopardy! questions, shows the potential current flexibility in discourse and ‘comprehension’.

    So a lot of this is a lot more possible than we might think.

  9. (If you’re not going to use BBQuote, at least allow posters to edit their malformed posts… [But I don’t think anything was lost. So, maybe not. :)}

  10. There is an article in MSDN Magazine related to this topic:

    Lowering Higher Education

    David Platt

    “I have a front-row seat to the coming revolution. One of the main services I provide to Microsoft and the developer community is bridging the divide between them and academia. As with newspapers, I expect the existing order in academia will be redefined over the next decade, posing enormous challenges and huge new market opportunities for the institutions and people who can grab them.”

    The rest continues here:

  11. runs a full catalogue of online economics/history/etc. courses. Take a look to see what you can incorporate.

  12. I just finished reading Seth Godin’s “Stop Stealing Dreams” e-book a few days ago, so my position is somewhat biased by that at the moment, but, shouldn’t we be celebrating the wide dissemination of basic knowledge like this?

    Yes, I understand the risk that some people will say “I watched all these videos, so now I know statistics,” but the more driven learner will find more videos, more books, more information, find other interested people, and put it to practice if they are truly interested.

    I really see no reason that curious person with a 21st century internet connection couldn’t become an “amateur expert” on any topic. We have a digital Lyceum right in front of us, and we still insist on churning students through in-person lectures, tests, and years of pointless schooling.

  13. Dear Professor Matt,

    I have been taking your online course for years now. I absolutely love it. Thank you a thousand times.

    If I miss a day, which sometimes happens, I go back and catch up. I think I have read every lesson you have written.

    Would you like some payment? I would be happy to chip in something. Just ask.

    This class may be the best one I have ever taken in my whole life. There’s the stats, but then there’s so much more.

    I’ve recruited quite a few students. They all thank me later. Really, they should be thanking you.

    Your faithful pupil,

    Uncle Mike

Leave a Comment

Your email address will not be published. Required fields are marked *