Discuss the CaMILE Project Here

CaMILE is a really a general term for a family of computer-supported collaborative learning environments developed by Mark Guzdial and a large number of collaborators (to mention a few, at least, Jennifer Turns, Cindy Hmelo, Janet Kolodner, Farrokh Mistree, Janet Allen, David Rosen, Jorge Vanegas, David Carlson, Noel Rappin, and many others). The original CaMILE (below) was a Macintosh-based collaboration tool. Unfortunately, it suffered from slow speed and being Mac-specific, which both limited access (most students, if they own one, have a Windows computer) and inhibited those with an anti-Mac religous fervor. Thus, WebCaMILE was developed (by David Carlson, based on the original CaMILE design), and is now entering its third generation, having spawned variations such as Web-SMILE.

Domains/Units

Domain is mostly irrelevant with CaMILE. Some domains may require notations or media that the Web doesn't well support yet (e.g., equation notation in math, or high-quality audio in music), but in general, any domain that can talk about its content on the Web can work well with CaMILE. CaMILE's focus is on getting students talking to one another.

Learning Problems

Originally, CaMILE was designed to support students in meeting when physical meetings were difficult or impossible. However, students were not really interested in using CaMILE for that -- the limitations of text-only messages made face-to-face meetings easier and more information-rich. The researchers did find that CaMILE excelled at getting a broad cross-section of students (typically, far more than half, which is more than might speak up in a lecture) to carry on sustained discussions.

WebCaMILE (pictured below) allowed for the use of anchored collaboration. Anchored collaboration, broadly speaking, is "giving students something to talk about". Specifically, a link could be created from any Web page to any CaMILE note, enabling teachers to create discussion spaces specific to given topics or pages. Anchored collaboration has been shown to be a huge benefit in getting students to sustain a discussion.

Kinds of Technology

CaMILE is a Web-based technology. Other networking possibilities (other than the Web) are feasible for CaMILE, but the Web provides a ubiquity of access which has a dramatic impact on easing barriers to student use. CD or other delivery mechanisms would not allow students to interact, a necessary component for collaboration software.

Development

CaMILE is a simple forms-based Web tool, originally developed with Perl, then later ported to Frontier, and now being ported to Squeak (using a toolkit called "pws"). There are limitations and strengths to this approach.

• Forms-based interfaces are weak: Simple text areas, buttons, and pull-down menus are all that you get. But they're ubiquitous and can run on anything. Better interfaces could be done with something like Java, but not all computers run Java and not all Java programs run on all computers.
• CaMILE was programmed from scratch. Commercial tools such as Lotus Notes have some similar features, but gave the researchers less control over the look-and-feel of the software. Plus, commercial software costs money, which is yet another barrier to student use.
• A challenge has been to get CaMILE to run reliably and quickly over the network. The original Perl implementation is mostly reliable, but is difficult to upgrade and change. The Frontier implementation had nice features, but could not run on any server and was too slow.

Integration

Since CaMILE mostly impacts student use outside the classroom, it can be very easy to integrate. One simply creates a CaMILE forum and informs students of its use. However, to use CaMILE effectively can require changes in the teacher's practice.

• Anchored collaboration doesn't just happen. The teacher must create Web pages with content of interest to students (like assignments or review pages), create a CaMILE note that encourages discussion, then link the Web page to the CaMILE note.
• The teacher (or the TAs, at least) should participate in the discussion. While a teacher does not have to participate, having the teacher participate shows the students that this is a valued activity.

Evaluation

Several evaluations of the effectiveness of CaMILE have been undertaken. Perhaps surprisingly, only a few of them have attempted to measure learning. While some evaluations have found evidence that students were learning from their discussions in CaMILE (especially the work by Jennifer Turns), trying to show class-wide learning from discussions is a challenging enterprise -- particularly across domains and across different teachers.

Instead, the focus of evaluations in CaMILE has been on mediating factors in learning: Are things happening which should be leading to better learning? In particular, evaluation of CaMILE has focussed on sustaining discussion. Do students talk together about a single topic for a reasonable length of time? On topics that relate to class learning?

The first study of this type did a head-to-head comparison of the same class in two successive quarters. Both quarters used essentially the same curriculum, the same textbook, and similar assignments. In the one quarter, the class used a newsgroup for class discussion. In the second quarter, the class used CaMILE. The teachers were different, though.

The results are summarized in the below two tables. In general, CaMILE discussion threads (a thread is a succession of notes on the same topic, commenting on one another) are significantly longer than newsgroups threads. Anchored threads (those that are pointed to by web pages) are much longer still. Further, the notes in CaMILE tend to be more on class learning topics, as opposed to questions about homework.

The second and more recent study looked at a large number of newsgroup and CaMILE-using classes, across domains and across teachers. (http://guzdial.cc.gatech.edu/papers/infoecol/) Again, CaMILE forums had longer sustained discussion than newsgroups across all domains and teachers. While this doesn't indicate that better learning occurred, it does suggest that better learning conditions occurred.

For More Information

• CaMILE home page: http://www.cc.gatech.edu/gvu/edtech/CaMILE.html

• Try out CaMILE (in the MMC form): http://guzdial.cc.gatech.edu/Tours/CaMILE.html