Slashdot Mirror
Software for Online Courses?
bcrowell asks: "I teach Physics at a community college, and a lot of the faculty are trying their hands at teaching internet courses. I'm going through the process of getting approval to offer some of our physics courses with online instruction plus lab, as an alternative to lecture plus lab. (My main motivation is to boost enrollment in some of our higher-level courses, which tend to get canceled if not enough people enroll.) The standard software for this kind of thing seems to be WebCT, but I get the impression that it's proprietary straightjacket-ware. I'd rather go with something open-source, especially since proprietary software seems to come and go, but the best open-source code is forever -- who wants to waste their time building a whole course around the flavor-of-the-month software? I'm particularly curious whether something like Slashcode might work. Most online courses include a requirement that people post a certain number of 'substantial' comments, where 'substantial' is a subjective term to be determined by the instructor. I know some teachers who say when they teach a large online course, they just don't have time to read all the posts, so they end up going by length a lot of the time. Wouldn't moderation by one's peers work better?"
Might there not be a zope product of some sort that would work for this? What are you looking for out of the project? What are your goals? "e-learning" or whatever seems a bit broad. Who will do the admin work? Does the university already have a site license for some other product?
For basic discussion type things, slashcode, or one of the clones/mutants might work rather well. In fact, too well. You'd need to modify it for closed registrations, drop the automatic mod priveleges, etc.
Finally, I can say that as a former student, I'd be least interested in a BS web-course at higher levels. At that point in my education, one-on-one time was highly important.
Jesus was all right but his disciples were thick and ordinary. -John Lennon
There's systems like blackboard.com for handling courses in a traditional lecture manner (one-to-many) covering a single topic.
We rolled our own for ThePhotoCourse.com using a combination of subforums and a way of doing lessons with integrated quizzes. This was a few-to-many approach: several instructors, each handling a small 'classroom' of similarly inclined students.
One professor has worked out integrated quizzes as an online component to his real-world course (which is where we got our approach for thephotocourse). The method is that you read the material and get quizzed, and if you don't pass the quiz you get hints or more material until you do have it 'mastered'. Still automated, but useful.
Ultimately, your course s/w is going to be based on:
a) your format: lecture/regurgative material, lab/hands-on, other
b) your teacher/student ratio
c) relevance of assignments and quizzes to the course
d) whether pass/fail (or grading) is required
Tools like slashcode are useful for _talking_ about something, but _teaching_ requires more than just two-way talking. You also need application, review, and testing.
Good luck!
A.
Well, grading on a curve is a barbaric custom. I can't figure out why anyone would do it. Actually, I was more worried about the opposite: people would form cliques that would always mod each other's comments up.
Even if there isn't a curve --- are you sure that having people who may not like you or necessarily have a good enough grasp of the material moderating your posts?
Good point, although a lot of what moderation gets used for on Slashdot is just getting rid of trolling and obnoxious behavior. That type of moderation doesn't even require that much of a deep understanding.
Find free books.
We are doing a "Pilot Program" (dear god in heaven..) of this company's main product. YMMV,but it seems to be a straightforward content-management type system geared specifically to class management (though not free, it is fairly cheap). I believe that if you bug them enough they will give you access to the source, but it isn't Free as in speech.
Developing class content for online presentation seems to me to be the stickier problem. To do it well, you either have to have been trained to be able to produce with a good tool like dreamweaver or authorware, or have access to a code-monkey to do it for you.
I had a sig, but
I teach philosophy courses online full time.
COURSE DESIGN: From the point of view of an instructor/course designer, WebCT is an extrmely flexible platform -- much more thn, say, Blackboard. If your one experience with the platform was negative, this probably reflects on poor course design on the part of your instructor.
INTENSIVE INTERACTION AND WORKLOAD: Students may sign up for an online course because they assume that it requires more work, but perceptive students quickly learn that the medium actually requires more time than a traditional classroom. The point is simple: the best educational advantage of distance education is the opportunity for much more intensive interactions between students and students and students and instructors. These interactions take time, but lead to strong educational outcomes. So, if course is designed to promote intensive interactions, the students will work harder and learn more.
Our office of about 3000 people has to take various short training programs annually...office stuff, sexual harrassment training, security, environmental policies etc.
When they moved it to the web they bought/wrote a package that does all the user testing using multiple choice questions and client side scripting. So rather than sit through an hour of online tedium I 1) view source, 2) identify answer 3) submit answer.
Nobody seems to notice or care that I'm done in less than 2 minutes.
...be sure to check out
Advanced Distributed Learning, the organization that promotes the SCORM standard for online content. SCORM stands for Shareable Content Object Reference Model, and is widely gaining acceptance in government and commercial settings as a standard for e-Learning courses. Tools which are SCORM-compliant have a better chance of becoming widely adopted, IMHO.
It is a GPLed course setup, developed and used by many universities accross the country. It stresses the "nonlinear" flow of a course, although it is easy to create linear "textbook" courses as well.
The entire loncapa is built upon Redhat 6.2 and is designed to scale my simply adding more servers.
Courses can be created within loncapa or they can be set up separetely and imported in. Each component of a course can be an HTML page, a multimedia clip, a Java applet, a problem or assignment. These components are combined into the course.
Keeping
Set yourself up with a webserver with httpd, irc, a maillist server, and a newserver. Then you have what is equivalent to webct. You can even get around the newserver by using a discussion board.
You will need to deal with security issues if you want your students to be able to check grades online. There are other security issues as well. I've been on a few maillists that got signed up to pr0n lists. This weren't webct lists though. Since webct uses a webmail type interface where only other members of the class can send mail to.
So, the question to ask yourself is: What will cost you more? Getting some monkey to set up standard software for you or shelling out some cash so that WebCT can set up proprietary software for you. Don't forget the maintainence costs.
If I were you and I'd have the IT department at your school set everything up for you. Use http for all the general stuff. Let people post questions to a discussion board. And hold tutorials over irc. Use the maillist for announcements.
You're looking at things entirely wrong.
The goal of a class shouldn't be to see which students are more intelligent than other students. The goal of a class it to impart a certain amount of information to a student. Depending on how much of that information a student absorbs, they earn a certain grade.
If all of the students in the class gain a solid and firm understanding of all of the material that they need to cover in the class, then hell yes they all deserve an 'A'. They've done what they came to do, and that's learn the material.
Using a curve can be acceptable in some (few) certain cases, in order to smooth over the differences between your expectations and your classes abilities, but things like bell curves have absolutely no place in education grading.
You mention that in the average class score is 35% then the test should be immediately curved so that everyone doesn't fail. I disagree. This means that one of two things has occured. Either the teacher overestimated the students abilities, or the students failed learn the minimum required information for that test. Either way, the situation requires more examination, and you shouldn't simply curve it so people pass.
Later, you say that an 'A' should mean that a student is in the top 10% of their class. That's idiocy. What if the class only has 5 people? What if they are all exceptionally smart, or exceptionally stupid? As a different teacher, or prospective employer, what do I care how someone compares to their classmates? All I want to know is whether or not they learned the material that the class covered.
To repeat, grades should be based entirely on how well a student learned the material that was presented in the class they are in. If they learned everything they needed to, regardless of if everyone in the class also did as well, or if no one else in the class did, they deserve an 'A'. If they learned almost all of it, they've earned a 'B'. If they attained a solid grasp of it, they should be given a 'C'. If they know the minimum amount that they need to of the subject, then they get a 'D'. If they haven't managed to learn the minimum information that the course requires them to learn, then they've failed, and get an 'F'.
If you're curious as to where I'm coming from, I'm a part-time college professor who's also still taking classes (and who prolly always will be). I've also worked for a number of years in the private (business) sector doing computer stuff. Thus, I've been on all sides, as student, teacher, and business.
Topher
You think blackboard is costly?!?!?! It's practically free compared to Docent, Saba, click2learn, IBM's thing, Sun's thing, everything else. Most learning software has per-user licenses with no realistic option of unlimited use. Blackboard is very attractive if you expect large numbers of people to take the course.
And, although it is not open source, it is written largely in Perl, and the source is available when you buy it. When I spoke to Blackboard they did not officially tolerate customers modifying the source, but the acknowledged that it went on pretty widely.
-----
I've spent the last 6 months looking at online learning technology. It's a mess. We ended up with Vuepoint (http://www.vuepoint.com/) because they have a tolerable unlimited user license fee, and they have both content creation/management and course deliver/management tools. Many companies offer only one or the other.
There are a couple of 'standards' in this area, namely AICC (old) and SCORM (current). The latter stands for Sharable Content Object Reference Model, and is theoretically a way of encapsulating bits of an online class to allow, well, reuse and sharing.
In practice the standard appears highly complex and unhelpful. It comes from the aviation world (don't ask) and is geared towards courses like 'how to troubleshoot a Bowing 747 hydraulic system'. I.E., big, fact based piles of knowledge. This makes it not very helpful for all sorts of other courses, like 'What to expect from chemotherapy'.
Nonetheless, in corporate world being SCORM compliant is a vital feature for many companies that want to sell elearning products. But, SCORM is a highest common factor, and everyone ends up sinking to that level if they go with it. There are companies that simply sell SCORM compliant classes for you to plug into your SCORM compliant LMS (learnign management system), but that's a pretty blunt approach.
The academic world is better - there are a number of products (like blackboard) that derive from universities' internal elearning projects. These products are often technically kind of a mess and pretty crufty, but they are often more functional and cheaper.
-----
I have led the implementation of WebCT for a major university and we have had MAJOR problems with WebCT (admin, usage, training, support, etc.). There have been a lot of changes in this market over the last year and to add to the problems WebCT has made some serious changes to their business model as they moved away from their university roots and into the corporate world (among other things this has resulted in a 500% change in pricing for campus licensing and a total rewrite of the Perl based version that can now be viewed as abandonware). Obviously if our students and teachers actually used WebCT we would have had to swallow these changes and continue on the bumpy road dictated by WebCT... but luckily that has not been the case.
For this reason we have been evaluating lots of different systems over the last 5 months and in our search we came across an application that is built on top of an industrial strength web application framework that was born in the CS department of MIT and has a very active open source community supporting it (http://openacs.org). All data is stored in a database (Oracle or Postgresql - i.e. it can be spit out in any standard compliant format needed in the future) and it is served using a heavy duty open source web/application server (AOLServer). The actual application that is built on this framework (http://dotlrn.mit.edu) is focused on community building and sounds like something that would probably interest you. We have decided as an institution to go with dotLRN (with more than 30,000 users), but we are still in the planning phases right now (we should be going live in the first or second quarter of 2003). Word of warning... although dotLRN has been in active use at various institutions around the world for more than two years (in an earlier version called AECS), the "boxed" second version that MIT is funding (dotLRN) is still in a testing phase (although it is being used in production at more than one adventurous school at the moment). Do not be upset though... the present version can be pulled from the public CVS and it works very well... and the "boxed" version should be coming in September. DotLRN was built with scalability, performance, community, and modularity in mind (i.e. modules that expand its present capabilities are being built as a write this: xml-based LMS, quiz module, presentation module, bookmarks module, glossary module... to name a few) and it is going to make some waves.
http://dotlrn.collaboraid.net
Greets!
Well, our system may be suitable, if you want to take a longer view *8-)
We're currently developing WHURLE - Web-based Hierarchical Universal Reactive Learning Environment - a GPL'd XML-based open system. It runs on a servlet engine with XSLT (apache tomcat and cocoon recommended *8-). The development team comprises several people who have run major online courses, mostly using WebCT, and we're keen to avoid the same mistakes - to this end - all our content is stored as discrete XML 'chunks' of information, which are structured into a hierarchical lesson plan - we're developing adaptive filters at the mo, but if you just want to do non-adaptive, straight course delivery, it's ready now.
I said take a longer view, because, even though we have courses running in Nottingham,UK and Hong Kong, authoring at the moment is either done by hand in an XML editor, or using some crude web forms I knocked up - we are in the process of finishing a converter for most of our legacy material, which was developed for a program called Scholar's Desktop - so if any one has courses built using this, let me know and I can get you started real quick - for others, if you'd like to see a demo of the system, or contribute, please drop me an email.Convertors from other formats are in hand, but realistically will take some time.
Addressing some of the other issues posted here - we think SCORM is a good idea and will get round to doing something with it RSN *8-) We found in our previous courses that what makes online components a success is a 'critical mass' of online contributors - silent forums are the death of any system. As for peer review - we're actively looking at this, perhaps using something along the lines that sourceforge use - most active, highest rated and so on, rather than the karma/mod system used here.
Hope this is interesting to some people - get in touch with me if you'd like to chat further.
Adam Moore
WHURLE Technical Lead
University of Nottingham
http://whurle.sf.net
The problem with the rat race is, even if you win, you're still a rat!