Universities worldwide are starting to offer online courses. Personnally, I just launched my first Computer Science graduate course (INF 6104, in French).
Because students can take the course on their own time at home, without having to show up at designated time, accessibility is unbeatable. (Note: some online courses still require you to show at designated times.) However, the transformation goes deeper than just having increased accessibility… the consequences are far reaching!
Some students need a professor holding their hand (see FSP’s recent account). These students are apparently reluctant to use Google, to seek the answers for themselves. In fact, they are probably perfectly able to browse, search, and read on their own. What they need is a strong emotional bound to peers and a teacher. These are the same students who will prefer to wait 30 minutes in line for a human teller instead of using an automated teller. These are the same people who will go to a crowded bookstore to buy a book, even if they have to drive 50 minutes to get there, instead of buying their book online in 3 minutes. What they seek is the human touch. Alas, in learning, shopping, traveling, the human touch will soon be an option. For computers to alleviate our work, we need to learn to rely more often on computers, and less often on human beings.
If you paid attention, you might see that I am drawing you somewhere… online teaching… automated tellers…
To a large extend, what I do with my students is to automate more and more of the teaching process. I no longer assign problems and then ask students to come to my office if they cannot do it. I post the solution online! I no longer wait for the unavoidable questions (when is homework 1 due?), I post all of the answers online. How do I know that I have posted all of the answers? Because 99% of all questions I get from students are answered on the Web site! I even get to tell my students about related stories in an automated manner (by using a blog). I work very hard so that I never need to repeat myself. I write it once for all students, for all times.
And I am not alone. My colleague Guy Tremblay designed a system called Oto. While I do not think he would spin it quite this way, the goal of Oto is to automate the marking of programming assignments.
A very popular service to learn English as a foreign language is Gymglish, an online automated teaching tool. Reviews are very good.
I conclude with some expected consequences:
- Automated teaching will make learning far more accessible and, eventually, slightly more affordable. The great benefit will be to be able to get a degree in nanomolecular physics from the North Pole.
- Students who do not need the “human touch” in any case, will get more learning done in a shorter time with automated teaching. Other students will complain that professors are no longer “there for them”. Automation will transform teaching, not make it better or worse.
- Just as computers are transforming the jobs of travel agents, bank tellers, and bookstore managers, they will have deep and profound consequences for teachers and professors. It is just a matter of time! Consider all the routine operations you do and see how you can outsource them to a computer! How much time do you spend in class telling students about when the next midterm is, and what is covered? Telling them how to get the last set of lecture notes?
- There will always be students who need the emotional bound with a professor and with their peers. Automation can only go so far. That is, until we can achieve true AI and people start becoming friend with their computers!
What is the essential role of computers or the Internet? Textbooks plus occasional teach-yourself bootstrap manuals are sufficient to learn pretty much anything as it is. If the problem is getting feedback, it looks more like a problem of creating high-quality thematic forums, this doesn’t fit in the format of a course, especially for standard stuff.
What is the essential role of computers or the Internet? Textbooks plus occasional teach-yourself bootstrap manuals are sufficient to learn pretty much anything as it is. If the problem is getting feedback, it looks more like a problem of creating high-quality thematic forums, this doesn’t fit in the format of a course, especially for standard stuff.
Textbooks have numerous limitations for learning, and the most serious one, compared to a university course, is that a textbook is not going to certify your learning. In some/most countries you cannot be an engineer, a doctor, a dentist, and so on, without taking the courses. What good does it do to learn dentistry on your own, if you can’t be a dentist?
A textbook is fine,but it does not come with a final exam and a set of homeworks. These must be updated regularly which is easy on the Web. In practice, textbooks are not always available or not always of the right level or quality. Many textbooks lack a large number of solved problems—these are supposed to be provided by the lecturer.
What about timely information? Textbooks are rarely updated every year… your 5-year old textbook on database theory is probably still good today, but it may not say much about the currently fashionable topics…
What about integration with Wikipedia? What about linking to published papers (PDF files)? What about integration of videos (YouTube, Google Video…)… Integration of other multimedia content (flickr)… integration with blogs on the topic…
For numerous purposes, online courses are far superior to a textbook… and again, merely owning a textbook is not going to get you a job.
Just as learning, certification could be extracted from the course format (and condensed to a necessary minimum). Why do you need teaching part at all, and if you do, why do it in a course format? (I meant textbook as an example, not as a bound on media.)
Vladimir: Teaching is an open-ended term. I teach Google Mail to recognize my spam. Life teaches you lesson.
Now. Do you need an individual teaching you something so that you can learn? No! That was the whole point of my post!
Jonathan:
being able to ask any questions you like
Students taking my Information Retrieval graduate course can ask any question they like. However, there are very few “original questions” that may arise and that nobody has ever thought up. Students who first learn some material are quite predictable. Heck! Even researchers are quite predictable in their questions!
Moreover, once I learn of such a question, I quickly integrate it to the material so that the next student will get the answer without having to ask.
Consider also that I link to a massive number of resources (research articles, wikipedia, and so on). These contain a lot of very good answers.
There are, yes, some questions that are better answered by a Ph.D. holder. And indeed, that is precisely one of the reasons I have a job. These questions include the resolution of contradictions. What if one reference you provide says something, and another says something else? A student might not be able to figure a way out.
ask the instructor to repeat themselves in a different way?
Sometimes students require a different formulation. It is provided, and once done, it is added to the content of the course, dynamically.
Typically, however, what the students really want is an example or two or three… once you have provided these, you are pretty much set.
Jonathan:
Maybe this works for certain subjects (basic programming comes to mind) but I can’t see it working well for more advanced ones.
Well, advanced topics require a more advanced instructors. It is best that graduate courses are taught by subject experts, preferably with some research experience in the field.
However, advanced topics work better online than basic ones. I would never try to teach kids how to read online… but teaching Theoretical Computer Science or Nuclear Physics online is comparatively much easier. In some cases, linking to the research articles might be enough if you have really advanced students…!
The same holds true for e-commerce. Basic shopping (milk, bread) works better offline… but fancy shopping (some computer part) works better online.
Are you seriously claiming that the only advantage of learning from a teacher vs. learning from a textbook or on-line materials is the “strong emotional bound to peers and a teacher”? How about the advantage of being able to ask any questions you like, or being able to ask the instructor to repeat themselves in a different way?
Maybe this works for certain subjects (basic programming comes to mind) but I can’t see it working well for more advanced ones.
Miguel: Online courses are no different evaluation-wise than traditional courses. The students does some homework and he hands it in. When applicable, he has to write an exam in some designated room at some designated time and he has to bring some ID.
E-commerce and automatic teller have not increased fraud. They bring about different types of cheating, and they may allow more sophisticated cheating techniques, but the problems are fundamentally the same.
Who is this person? Did this person do the work? How do I know that this ID is legitimate? What proof do I have that this person really master the material?
These are hard questions.
University professors, at least in research institutions, do not have much time to hunt down cheaters. So cheating will happen. Online or not.
In this context, how do you evaluate your students and (most important) how do you prevent fraud?
I like the article, but I think you’re using the wrong comparison; online classes should be compared to online jobs rather than teller or bookstore transactions. I believe that online jobs are still a limited development for the same reasons as online classes. There is a lot more human interaction that goes on at a typical workplace or classroom than at a typical teller window or bookstore counter.
the education industry will hate and hampper every effort to establish on-line education ..