Sakai 3.0 Proposal Comments - Extended Version

These comments are in response to the Sakai 3.0 Proposal at the Sakaiproject. A shorter version was posted to the Teaching and Learning email list on the Sakai Collab site. The original version is on another page.

Background

First, my background with Sakai and online learning management systems: I support instructors at the Stanford Language Center, which covers foreign language instruction for undergraduates and ESL for graduate students. I also teach in the ESL program for graduate students (English for Foreign Students), and have used Sakai since we began piloting it in spring of 2006 as CourseWork 5.0. My familiarity with online course learning management systems also extends to using it as a graduate student, where I used CourseWork 3.0 (aka CourseWork Classic) , in courses for a grade. However, my responsibilities now allow me to have super-user permissions on our current version of Sakai (2.4), and I participate in many of the support and planning meetings.

New possibilities

Overall, I found the possibilities raised in the Sakai 3 Vision proposal to be quite exciting. The proposal to strengthen Sakai as an academic collaboration tool is very encouraging, since online tools have so much to offer in this area. Improving the content creation ability is also a welcome addition to the proposal, especially because the idea of a “flexible widget-based user experience” reflects current trends in consumer-based operating systems. Also, the move to look “beyond sites” will add much needed flexibility to Sakai, perhaps even in areas beyond those discussed in the proposal.

However, while these are all very important potential improvements, I felt that perhaps there might be other areas that deserve consideration. There is a need for more focus on the enhancements to one of the tasks that Sakai already does rather well: managing courses. There is huge potential for improvement in the way that Sakai can support teachers, beyond student-teacher and peer relationships, but I see nothing in the vision proposal related to these areas. For example, it addresses how awkward content creation is, but does not mention how this varies by role. Students (in “access” role) may spend many hours working in Sakai, but it is usually across several courses, while instructors (in “manage” role) must plan courses and lessons, and then assess the range of students in the class. User interface issues are immediately apparent: most student-level pages will fit on one screen, while instructors usually have to scroll to see a full page, often with the Save button at the bottom.

Roles and Permissions

However, beyond user interface issues, Sakai seems to be designed with only a two-dimensional teacher-student interaction in mind, especially when looking at the roles and possible permissions. The proposal’s emphasis on collaboration addresses peer level interactions, but there is no mention of any possibility for establishing hierarchy. However, hierarchy is a reality of education around the world: teachers are often grouped together into units such as grades and departments with administrators such as coordinators and department heads. Curriculum is mandated from above, or shared among instructors, but often delivered to students in a very different form. In the department where I work, we have tried to create a set of permissions that would enable a coordinator, who is in charge of a group of instructors, but we ended up just creating new “resource” sites just for instructors to get mandated material and assignments. Even then, however, the roles are not what we would like: If the instructors are made students in that site they can download documents, but cannot export assignments. Consequently, they are left as instructors, but coordinators have found cases where instructors have accidentally deleted material. Also, coordinators create certain required assignments that instructors import from these resource sites and use for the term, but if they find there is a mistake, they need to have instructors download them again. Ideally, however, if the author makes a change, it should propagate out to all copies. This has been implemented for assignments within a site here, after considerable planning and effort by the developers, but there is no concept of a hierarchy beyond the site.

Concepts of facilitating instruction

The vision proposal also does not show an understanding of the different areas where teacher-student interactions happen. Sakai seems to be almost entirely skewed toward facilitating homework such as assignments and reading materials, with very little in the way of facilitating a 70 minute classroom lesson. If our concept of a 70 minute lesson is simply a lecture where students take notes, then, indeed, there certainly is not much to facilitate, but if we look at the types of activities that expert teachers might make use of in a lesson, we can see many more possibilities. Small groups are almost always part of these activities, so it seems that the ability to quickly make ad-hoc groups of students, and provide each one with unique, permission specific resources would be a good place to start.

One of the tests I use for any application of classroom systems is a listening activity for language learning that I have worked with for many years. In the low-tech version, I prepare two short (30 sec) audio clips of news or some other natural speech, along with a transcript for each. I put students into two groups and give each group one of the transcripts and a pair of scissors and tell them to cut out 15 words to make a fill-in-the-blanks listening exercise for the other group. When they are done, they exchange cut-up passages and listen to each passage once or twice (as a group) and try to fill in the blanks (as a group). The object of the activity is for them to try to figure out what is difficult to hear in natural speech (function words and reduced forms). Translating this to a high-tech version seems like it would very straightforward, but I have run into several difficulties. Using the resources or assignments on the Sakai site is ruled out from the beginning, because any documents uploaded are visible to all members. The drop-box is a good (almost low-tech) way to do it, but it requires putting a copy of the document into each and every student’s drop box, keeping track of who is in what group. Then the “switch” requires distributing the group-created problems one by one again. Also, with this activity, as is often the case with in-class work, there is always very little for the individual students to go back to if they want to review. Moving to a high-tech implementation should accomplish this, but in reality, is not straightforward. Finally, it is important to note that instructor control is essential: Once the playback portion starts, allowing only one pass at a time through the audio really emphasizes the characteristics of natural speech that make it difficult. Unfortunately, while a teacher can control this trivially with a cassette player in front of a class, it is almost impossible to control over the on computers with individual users. A couple years ago, I did something similar to the drop-box option in our language lab, but found that the activity was far less communicative than the low-tech version, because each student was on a separate computer, reading the transcript and listening to the audio file with their headsets on. This experience showed me the individual nature of student-computer interactions, which runs against the communicative aims of the language classroom. The solitude starts with the first login, which leads to a private page, but could be reduced by somehow providing the resource to a page that is shared by all members of that ad hoc group, and viewed on a single, shared screen – this is essentially a hardware requirement and is perhaps beyond the scope of Sakai.

Creating and breaking boundaries

Also, I found myself somewhat perplexed at the inclusion of social networking among the primary driving forces in designing the next generation of Sakai. If this is meant to be an effort to enable RSS feeds to be sent out and pulled in, then I can see many benefits if authentication and privacy issues are addressed. However, many institutions use Sakai as a course management system, and therefore rely heavily on the ability to implement authentication so that students can’t see what teachers do until they need to and the public can’t ever see what students do. Social networking is very exciting and useful, but it seems that Sakai would be better going with its strengths, rather than trying to do something that Facebook, MySpace and LinkedIn will always do better. Sakai does a good job of facilitating course management, and can offer the benefits of working with an open community, but institutions need to retain and guarantee control of privacy and of resources in their own implementations.

Finally, there is also mention of open teaching practices. While this may be a worthy initiative to contribute to education on a global scale, I think that many teachers are not very comfortable with the idea of their materials being available to an audience beyond the students enrolled in their course. The lessons they create are the product of all of their education, work and experience and some teachers are very hesitant to just give that away. Just as facilitating classroom interactions would be an area for Sakai to improve, the ability to facilitate distance learning is another important area to develop, but it is difficult to imagine how a conscientious teacher could commit to providing a useful level of interaction with an unspecified number of students.

Overall Direction

I feel strongly that we should not miss this opportunity to use Sakai 3.0 to re-think the LMS and academic technology in general, and this will involve serious consideration of what Clay Fenlason termed, the “user-facing functionality”. Maybe the things that LMSs do were revolutionary 10 years ago (15?), but now just about anything online can upload documents, facilitate collaboration and provide some boundaries. I teach a series of classes for non-matriculated students (in Continuing Studies) so I can’t use our LMS, but I use 2 pbwikis (one for administration, one for collaboration, for a total of 4GB of storage), grandcentral and email and accomplish everything I need. In fact, grandcentral offers my students a telephone interface for an audio response, which is something that Sakai hasn’t even thought about yet.

Other ideas

It is very valuable to talk about new ideas like social networking, but there are any number of other things that could be included in future versions Sakai, whether that is 3.0 or later. Just to name a few:

Mobile
All the end-of-the-year Future of Technology articles / blogposts mention the growth of mobile applications. We have a rather surprising number of students with iPhones here on campus, and, more importantly, the percentages are the highest of all in the freshman class: the numbers aren’t going to do anything but go up.
Posting video
And I don’t mean just a way to embed a YouTube clip. Students are more and more comfortable creating video, and some courses here are accepting (even requiring) multimedia presentations for assignments. Video files are usually huge things that do not do well uploading and downloading, especially for teachers who live off campus and don’t always have smokin’ fast internet connections. Why can’t we have something that automatically converts video files to streaming? Why can’t there be something like YouTube in Sakai? Flash Media Server includes instructions for building something like this for free (almost) in their sample applications. The thing that is missing is the fine-grained permissions, so that I can limit the viewers to just me and one student, or share it with the whole class, or share it with the whole university.
Watching video
There is more than just listening going on. I saw a presentation here about some technology at scpd.stanford.edu that uses time-stamps to allow the viewers to collaborate on note-taking, question-raising and answer-proposing.
Testing
Go beyond the usual multiple choice / short answer / file upload options. In the languages we follow ACTFL proficiency guidelines which are built around an Oral Proficiency Interview. These are one-on-one sessions that can last 30 minutes, so assessing each individual student can be prohibitively time consuming. So for the last 7 years we have done a Simulated Oral Proficiency Interview that interfaced with CourseWork with custom built software. The next version will interface with Sakai and once again we have been struck by how difficult this is to do: Media must be controlled so that it plays once and once only, recording must start and stop automatically with no opportunity to pause, and the interview must move on to the next item automatically with no chance for going back. Simulating a conversation like this is very difficult to do on the web, but something that teachers need to do all the time.
Grades
And while we are on testing … why not put some item analysis in there: classical and maybe even some item response theory number crunching. Maybe even include pseudo-random draw based on the performance on the previous question. Tools for figuring out a curve, custom graphing to really show students the point spread and arbitrary curve on last week’s physics test, comparisons to last year’s items – yes, items, not assignments. This is what computers are good at, isn’t it? We’ve got the databases right? This involves an extended concept of question pools, which might be another candidate for open education.
Privacy
Helping faculty and staff comply with data security regulations. In a sense, universities have been doing cloud computing for years, with unix systems. Learning management systems are quite similar in that students can access them anywhere and securely store their data. For this reason, I tell the instructors I work with that our LMS is the best way they can avoid FERPA violations: If they just leave student work, rosters and grades on the LMS, then it is not on their laptops and there is less to worry about if those go missing or get stolen. Are there other ways that we can use this? Can we work with the various IRBs to make sure that the LMS can be a place to store human subjects data?
Research data collection
Start with psychology. Researchers use software to collect data from one subject at a time: sets of responses, reaction time, etc. Wouldn’t it be cool if they could do that with a whole classroom of subjects? Wouldn’t it be cool if the experiment was a part of a pedagogically valid activity? At the very least, make data collection via a flash application (maybe in the SCORM player?) possible, and you will get a whole new group of people interested in using LMSs.
Indexing and search
I am at a disadvantage here, because the version of Sakai we are using at Stanford does not have any such capabilities (to speak of). So maybe a later version has it, but how efficiently can you filter? Can I find all documents by a certain student? Can I find all instances of a certain word? How about logging? Can we find which departments have been posting more materials over the last year? Can we compare student or instructor time on task over time between departments and even assignments? If we want to justify budgets for investing in Sakai, then we need to provide detailed numbers showing that it is both being used and being used effectively.
In-class activities
This is what the above language learning example is about. Yes, it is very specific, but the issue is that we need to think about what people do in classrooms, as separate from homework, as separate from distance learning. What can technology do for us in the classroom beyond projecting PowerPoint?
Self-study
This is where the open education topic has application for me. It has been useful for me to start to think about activities that students can do on their own as opposed to activities that I would lead as a teacher. With the 24/7 availability of the web, I have realized that I can and should provide my students with material that they can study on their own. In the past, this used to be part of the class, and teachers would use it when they had time. But if we put it online, then the more ambitious students can take advantage of it whenever they want. However, this should be combined with some sort of learner training for all students, to make sure that they know how to efficiently use the resources available to them. This learner training is not yet fully explored, but I am working on a colleague here with some techniques for language learners (come see us at CALICO).
Interoperability
Can this connect to other applications? Is there some way to have it interface with Matlab, or whatever the science students use? How about the design and debugging software that the EE or CS students use? Can linguists draw their tree diagrams somehow? What would even be a useful form for instructors in these departments to get the work their students produce? And is there a way to facilitate collaboration between professors who don’t just create text documents?

Focus on the users

Maybe these are already being seriously considered for Sakai, or maybe they are specific to my field, but my point is not the details but the direction. I would argue that the one of the most important tasks in front of us is to start surveying various fields and trying to think seriously about how technology can really change the way that we teach and learn. The LMS of tomorrow should not just be a replacement for what has been done on paper, but a way to facilitate and even foster the innovation of teachers. I cannot stress this enough: the goal is not just to facilitate instruction, but to facilitate innovation.

Finally, as Ray Davies pointed out, it is important to make it clear who is jumping to whose tune. One important implication in the Sakai 3.0 proposal was usability. However, I think this should be expanded so that usability really has an established place in the Sakai project. I had the opportunity to talk with a user experience engineer at Google who explained why their products are just so easy to use: they run every imaginable user test on them, including eye-tracking and at-home observations. We can’t really expect to be able to do that with Sakai, but if we are honest, our users have come to expect that kind of usability from anything they use online. Some sort of standards should be set out for testing the kernel and tools before they reach schools that will depend on them. So perhaps I am saying that that the user experience side should call the tune, if for no other reason than the fact that the people who decide the budgets to give to Sakai related projects are usually strongly influenced by user-facing functionality.

 

January 15, 2009

 

original version

 

ken's top page