Learning Tools | Special Feature

A Guide to K-12 Open Source LMS Options

The key decision when selecting a Learning Management System (LMS) for a K-12 school district is no longer whether to use a commercial option or the open source Moodle. Instead, costly commercial solutions are increasingly left out of the equation, and administrators are instead focusing their energies on which of several open source systems might be best for their district.

When conducting needs assessments for LMSes, administrators are looking at a variety of options and coming to different conclusions. While Moodle and Sakai are the two open source LMSes with the biggest market share in the United States, there are more specialized and newer entrants to the field that are getting notice, including Instructure's Canvas, OLAT, ATutor, and Google's CloudCourse.

In Maryland, educators conducted a full needs analysis and comparison of open source LMS options with funding from an EETT partnership grant. Eleven of 24 K-12 school districts in the state participated in the process. Of those, about half were using commercial products at the start of the exercise, while the others primarily were using Moodle.

Christine Voelker, grant manager for the Open Source LMS Partnership Grant (OSLMS) and based with Calvert County Public Schools, said that the project launched because districts were starting to consider open source solutions.

"The commercial products were getting more and more expensive, and a lot of restrictions were placed on what you could do with them," she said. "Across the state, we saw people were interested in open source, but they were hesitant because they didn't know much about it and what support was needed."

Through the OSLMS project, organizers selected three options--Moodle, Sakai, and SharePoint SLK. They then created a functionality rubric, which was sent to each of the partnering school districts, to evaluate the options based on three roles--system administrator, facilitator, and learner. Simultaneously, they set up virtual environments for each.

A basic course package was developed, and reviewers from across Maryland were hired to serve as reviewers in the three roles. The reviewers attended a webinar training program that Voelker conducted on how to access each LMS, work with the documents, and complete the functionality rubric. Each reviewer completed the tasks on the functionality rubric within each LMS and gave scores. 

The choice came down to a vote between Sakai and Moodle. "The data for these two LMSes was extremely close, whereas we could rule out SharePoint SLK easily," Voelker said. Sharepoint SLK was too bare bones, she said, and they had trouble getting one component to work at that time.

In the end, Moodle won the vote, largely because some of the school systems had already started with that system, and they were able to receive training and support through the grant. Because of this, they also were able to compile data on self-hosted and hosted environments. (Calvert's SaaS implementation of Moodle is hosted by Moodlerooms, a Moodle service provider and developer. Moodlerooms also provides a non-open source, commercial LMS called joule that's derived from Moodle, but a spokesman said that Calvert's implementation is open source.)

The next step in the OSLMS process is to develop a Total Cost of Ownership (TCO) worksheet to help districts plug in their own data and get a sense of how much an open source system would cost. "School systems will be able to then look at this in comparison to the costs of using their current commercial LMS and determine if it would be cost saving," Voelker said.

Does this process indicate that districts in the state might ultimately implement the same LMS?  "I don't know whether we'll have a state-wide solution," she said. "That would be wonderful and something we'd really like, but each local school system is on its own."

Moodle has been evolving with its current architecture since 2001. The current version, 2.1, was released in July 2011. The software has been translated into 82 languages.

In New York and Ohio, K-12 administrators also conducted analysis of open source learning management systems but selected a different option--Sakai.

Ryan Veety, network security analyst for the Minisink Valley Central School District in Slate Hill, NY, started looking at open source alternatives to Blackboard in 2007. The instigating factor was cost; as more teachers became interested in online learning with their students, he said, Blackboard licensing fees kept rising.

Veety installed Sakai and Moodle and began testing. He found that it was easier to install Moodle, which took only minutes to be up and running, while Sakai was significantly more work, requiring a Tomcat environment and building source code.

Despite that challenge, the district still decided to go with Sakai. "When comparing both Moodle and Sakai side by side, it was clear that Sakai was more user friendly," Veety said. "The user interface is simple and uncluttered."

In Veety's experience, Sakai had tools that technologically advanced teachers were asking for and also was simple enough for novice users to build sites for their classes. The district now has more than 400 sites in Sakai, and he regularly sees 50 to 100 simultaneous logged in users throughout the day.

Thankfully, Veety said, there is good documentation on the Sakai Project Web site, along with active user and developer communities. "The performance is Sakai is excellent," he said. "It is clearly designed for institutions much larger than ours."
Use of Sakai has resulted in significant cost savings for the district since managing a Sakai server is no more difficult than managing a server running another LMS. "The last time the server was even rebooted was 379 days ago, so I would say the technical support cost for me is near zero," Veety said. "Last summer, I upgraded to 2.7.1 and I haven't touched it since.  This summer I will upgrade to 2.8."

Paul Ross, technology director of Bexley City School District in Bexley, Ohio, started with a desk audit that examined criteria like user base, support options, roadmap, interoperability, scalability, funding base, development community, user interface design, and viability in the marketplace. From that, three options were selected: Moodle, OLAT, and Sakai.

OLAT was the first to be eliminated, largely because they had a small user base outside Europe, and at that time, it seemed development had stalled, he said. The district ran a pilot for six months on Sakai and Moodle, with 250 students and 16 teachers evenly distributed across the two platforms and across the five schools in the district, from grade four to 12. Students and faculty members submitted bi-weekly surveys.

Results showed that teachers favored Moodle slightly more than Sakai. "This in part was attributed to previous experience with Moodle," Ross. "Other schools locally were using Moodle, and there was a regular diet of Moodle training across the state."

The students had a different preference: Sakai. When asked whether they would recommend the LMS to others, Moodle was rated at 57 percent, and Sakai registered at 80 percent.

Ross said the final step in the review process was to examine which LMS was most in use at universities where district graduates were studying. In reviewing the previous year's 46 universities and colleges, they found half were using Blackboard or WebCT, a third Sakai, and the remaining using a variety of portal-based products. Moodle was not in use by any of the universities. 

That was a deciding factor, he said, because students want to use a platform found at universities. "It demonstrates that what they learn with us can be transferred to the next chapter in their career," Ross said. "There are many avenues for K-12 and higher education to share innovations, and Sakai is one example."

Teachers in his district have become passionate advocates for Sakai. "The testing aspects of Sakai are rather dynamic, enabling our teachers to use the system for formative and summative assessments," Ross said. "Sakai is a critical component of our instructional information system."

The system has been in place for two years, and about a third of the district is Sakai, though Ross said he expects a dramatic increase in usage the next academic year as they migrate use away from other paid applications and services to Sakai.
The district has experienced some growing pains around the occasional system error, he said, but those challenges have been overcome through a training program and planning for updates twice a year--not coinciding with class time.

The next phase, as the district enters year 3, is to engage a developer consultant to assist the district with integrating the platform with their student information system and creating a parent role in the learning platform.

The cost savings of using an open source LMS has been significant for the district. "Over two years, we've spent $0 on hardware or software, and training has been incorporated in our current program of professional development," Ross said.

Version 1.0 of Sakai was released in March 2005 after Indiana University, University of Michigan, MIT, and Stanford University--all of whom collaborated on its initial development--received a $2.4 million grant from the Andrew W. Mellon Foundation.

OLAT is another open source LMS option that K-12 school districts have considered, but have not selected, as far as THE Journal could determine. Developers at the University of Zurich--the home base for OLAT--noted that, historically, the focus of OLAT has been on universities and other higher education institutions, but that might be changing.

Joël Fisler, at the Multimedia & E-Learning Services (MELS) department at the University of Zurich, said that while OLAT is primarily popular in higher education, there has been enough interest that he held an OLAT workshop for K-12 teachers in June.

"We don't really know how widely used it is in schools," he said. "All our cooperation is with universities and other higher education institutions. Schools usually don't come to us, but to one of our commercial partners, or they decide to download and install OLAT directly themselves."

He noted that there is a difference in complexity between the LMS needs of K-12 schools and universities. "In OLAT, you have the possibility to create e-portfolios, define complex workflows for tasks, topic assignment workflows for master thesis and other thesis, and more," Fisler said. "So, OLAT includes tools and functions that would never be used in a K-12 school. In a school you want simple tools without lots of configuration possibilities."

Another aspect is navigation. Users can personalize OLAT courses as fine-grained as they want, but in a K-12 school you want simple navigation that preferably looks the same for every pupil and for every course or class. "You want easy configuration like, 'My online course has a folder: yes/no? my online course has a blog: yes/no?' etc," he said.

A third aspect is time schedule. In OLAT, users can have courses that run for a semester, for a year, or for multiple years. "Courses in OLAT are not dependent on a certain class or time schedule or time period," he said. "This would not make sense in a university context, but a school you want a tool that reflects and includes the weekly time schedule."

Development of OLAT started in 1999 at the University of Zurich in Switzerland, and the university continues to lead development.

ATutor, which emphasizes accessibility features for visually impaired and disabled learner, also is suitable for educational use, according to software evaluation criteria established by the American Society for Training and Development (ASTD).

Greg Gay, project manager at Inclusive Design Research Centre (IDRC) at Ontario College of Art & Design (OCAD), said, "We don't keep stats on K-12 users, or other levels for that matter. Only general stats like downloads, and registered installations, etc." However, he noted, ATutor features would be relevant for K-12.

ATutor was created by developers at the Adaptive Technology Resource Centre at the University of Toronto in 2002. Funding has come from the ATRC, the Learning Disabilities Resource Community, and the government of Ontario, among others.

The Newer Players: Instructure Canvas and Google CloudCourse
Newer players that K-12 districts are starting to take note of include Instructure's Canvas and Google's CloudCourse.

Google CloudCourse
Google announced its open source "course scheduling system," CloudCourse, in spring 2010. Built on App Engine, CloudCourse allows anyone to create and track learning activities and also offers calendaring, waitlist management and approval features. Early reviews, however, noted that it did not yet contain all the functionality that would make it an alternative to the other available LMSes.

According to a software engineer at Google who was quoted on the Desdemona blog in May 2010, the system was designed as a course scheduling tool for enterprises, but could be of use by school administrators, who most likely do not have the time or resources to worry about hardware hosting and dealing with traffic bursts.

On the other hand, John McLear noted in his School Technology blog that CloudCourse is not relevant for primary schools--and just because something is a Web-based tool that can be used within a learning platform doesn't mean it is a learning platform.

Still, it's something that administrators will look at going forward, as they implement a LMS for the first time or reevaluate their previous selection.

Instructure Canvas
Instructure Canvas is a new approach to open source learning management systems, using a software-as-a-service/cloud model of delivery. It has an open source version but is supported by a core company.

Mike Kisow, a Park City Schools' district technology instructional coach in Park City, UT, is one of the early adopters of Canvas, though he did not select the open source version.

The district had some experience with Moodle but was largely unsuccessful in attracting and retaining teachers as Moodle users, Kisow said. The push for a new solution came in response to the district's 1:1 laptop imitative.

"We were on an accelerated timeline as our first 1:1 school was less than a year away," Kisow said. "We utilized an already established [high-access taskforce] committee [comprising] administrators, technology staff, and educators."

After looking at a variety of options "at a distance," they scheduled in-person demos with Pearson, Blackboard, Instructure, and Aglix in March 2010. From this, they saw that Canvas fit their particular needs with its usability and its calendar, notification system, and file storage structure. Other major selling points included Canvas's mechanisms for making teacher feedback a "conversation."

Kisow said with the district's timeline and uncertainty with their first foray into 1:1, they wanted a stable hosted LMS solution. "Open source was not off the table, but our plates were rather full as it was," he said.

While Canvas was designed as a solution for higher education, it has translated very well to K-12, he said. Of course, some features could be improved, he noted, such as better support for block schedules and the large simultaneous courseloads typical of K-12 teachers and students.

"Canvas is one-stop-shopping for our students to quickly see the big picture of their educational careers," Kisow said. "Gone are the days of students needing to negotiate a tangled Web of class blogs, wikis,... teacher Web sites, and other random resources which amounted to a disjointed online coursework experience."

While Canvas is new, he said, that there is an active listserv, as well as budding wikis, and several users, including Kisow, have made their video tutorials public.

Institutions can use Canvas two ways. Instructure can run it in the cloud for them (using Amazon's cloud service); or they can download the open source code from Github and run it themselves. According to an Instructure spokesperson, running Canvas in the cloud is especially attractive to K-12 because they often own less infrastructure and hardware than post-secondary institutions.

Running the service in the cloud allows for a lot of flexibility, according to the spokesperson. For example, when a teacher changes the date of an assignment, a notification is sent automatically to each student in a way he or she prefers, whether by text message, Facebook notification, or e-mail.

By going with the open source version, institutions would not get migration tools to help move content from previous LMS, and the open source version does not include mobile apps, such as SpeedGrader on the iPad. In addition, most significantly, the open source version requires hardware to run the service and a team to manage its deployment--as opposed to the cloud offering where Instructure takes care of hosting.

"We were looking to test LMSes that had a proven track record with some significant history," said Voelker, from the Open Source LMS Partnership Grant (OSLMS). "I think that had Google's LMS and Canvas been available earlier, proving themselves to be formidable competitors that would stand the test of time, there is a good chance that they would have made it onto our short list."

Editor's note: We previously noted that the facts in our section on Moodle were in question. It turns out the article was accurate as originally published. We added a clarification stating that Calvert County's implementation of Moodle is hosted. [Last updated July 28, 2011 at 1:30 p.m.] --David Nagel

About the Author

Natasha Wanchek is a technology writer based in New York.