Bringing That CanDo Spirit

An educator's determined effort to update his school's archaic data systems results in a grassroots programming project whose success is all in its name.

Bringing That CanDo SpiritSOMETIMES THE BEST WAY to get something done is to do it yourself. Other times, it's the only way.

That's the conclusion David Welsh, an instructor in the Television and Multimedia Production program at Arlington Career Center (ACC) in Arlington, VA, came to in 2004 when he wanted to find a better way to track the progress of his high school students against state requirements. His search ultimately led to the creation of a successful open source application developed entirely in-house.

ACC, an Arlington Public Schools program that offers students the opportunity to learn specialized skills in areas such as aviation technology and culinary arts in addition to core subjects, had been tracking students' competencies via an outmoded pencil-and-paper method that Welsh believed was neither effective nor beneficial.

"We knew what the standards were, but there was no way to report against those standards," he says. "There needed to be a certain level of transparency so that not only would we as instructors know where the students were as far as meeting the competency level, but the students and their parents would know as well."

Welsh went looking for something that would keep a database of state requirements for each program at ACC, match the students' progress against those requirements, and create the yearly report that is sent off to the Virginia Department of Education. What he found, however, left much to be desired. "There was really no software that did what we needed it to do, and the only thing remotely similar was pretty expensive," Welsh recalls. "So we said, 'This may be something we want to consider writing ourselves.'"

Welsh recruited a colleague, Jeff Elkner, then a computer science teacher at Arlington's Yorktown High School, to create a program to meet ACC's needs. They had met a few years earlier at a professional development course in computer networking and had collaborated on an earlier project. "We had a great deal in common in the ways we were teaching and using technology," Welsh says.

Elkner zeroed in on what was lacking. "The system the state was using was basically a Word document that instructors would download and fill out," he says. "But even if you did all that, there was no way to do analysis of the information. I told [Welsh] that I thought it cried out for a web application."

Elkner turned to one of his best students to write the application. The student, Paul Carduner, wrote a rudimentary first version, which only Welsh used, but it demonstrated that creating an open source app in-house was the way to go.

"We really wanted to leverage our [district's] computer science teachers and students, and approach this project from a grassroots level," Welsh says. "When we started thinking about a solution that could run statewide, the solution we adopted had to be free or lowcost.... The state was already open to [using] open source software."

"Not only did they solve a real-world problem, but they did it in a way that was the best learning experience they could have ever had. These students left high school as professional developers."

Over the next two years, the application-- dubbed CanDo by Elkner because "it records what students can do"-- grew as a project assigned to Elkner's advanced computer science students, who wrote and finessed the code. In 2005, a chance meeting between Elkner and Tom Hoffman, director of the open source SchoolTool student information software, gave the project some much-needed support.

The two men met at PyCon, the annual conference for developers who use Python, the programming language Elkner was using to create CanDo. "[Elkner] was working on planning what CanDo was going to be," Hoffman recalls. "I had just started managing SchoolTool and I said [CanDo] would be a perfect case for SchoolTool," which was just getting off the ground.

The SchoolTool software suite features a turnkey student information system including demographics, a gradebook, attendance tracking, calendaring, and reporting for primary and secondary schools, as well as a framework for building customized applications and configurations for individual schools and states. Hoffman and the developers of SchoolTool offered Elkner and his students their expertise on how to integrate CanDo into their platform. "[Elkner] jumped right in, and both programs grew side by side," Hoffman says.

By 2006, about 25 of the district's computer science students were working on the CanDo project. Recognizing value as CanDo was fleshed out into a full-fledged operational program, Arlington Public Schools and the state of Virginia each awarded Elkner and his students $100,000 toward its development, enabling students to be paid for their work.

In 2007, ACC hosted a paid summer internship program for 12 students from Thomas Jefferson High School for Science and Technology in Fairfax County and 12 from ACC. The students began the internship during the previous December, taking programming language classes, and then met once weekly through the end of the school year and daily during the summer months to work on the application. "Out of that, we got four students who have since done ongoing development on CanDo and two who are total pros," Elkner says.

The students succeeded in developing CanDo into a web-based application that can be tailored to fit teachers' curricular needs. Teachers can use the app to create a list of 100 state-mandated standards of learning (SOLs) relating to career and technical education, with which they can track student progress throughout the school year. Meanwhile, students are able to monitor their progress online, working at their own pace. The application also allows teachers to upload documents and PowerPoint presentations that relate to a particular SOL, and includes a journaling feature for both teachers and students.

By 2008, CanDo was ready to be pilot-tested throughout Virginia, in 250 classes in 10 schools. And this past spring a second pilot was begun in another eight schools in the state-- one in every superintendent's region. "The feedback was good," Welsh says. "Ninety percent of the users said they would like to see it used statewide."

While pilot projects continue across the state, all Arlington teachers are now required to use CanDo. So far, according to Welsh, the software seems to be achieving the goals he set out to address. "[It] has brought a level of consciousness to teachers of those skills students need to achieve competency in the state's eyes," he says. "From a reporting aspect it has made teaching more rigorous. We teach hands-on projects that are exciting, but until you can define which competencies are associated with which projects, it's hard to see which students are performing at the standards. This software forces you to see which students can do the work."

Another important outcome of the CanDo project is the growth of the students involved in the application's development. "We basically had to train our own [student] workforce for CanDo, and it was an incredible experience for them," says Elkner, who last year joined Welsh on the staff at ACC. "Not only did they solve a real-world problem, but they did it in a way that was the best learning experience they could have ever had. These students left high school as professional developers."

But, despite CanDo's educational value and the app's availability as open source, tight school budgets across the state are preventing widespread adoption, at least for now. "Free software doesn't necessarily mean it's free," Elkner says. "Any program that involves personnel training and time requires a budget. And right now the money isn't there."

Still, hopes are high for the future of CanDo, especially given its fairly low cost compared to proprietary programs. "I have watched other states put out proposals for similar software, and the proposals come in at $1.2 million," Welsh says. "We came in at $200,000. It was built at a fraction of the cost, we paid the students to build it, and we're not at the mercy of the vendor because we created it and it's open source."

Welsh speaks of the use of students in the development of the software with a swell of pride. "When we apply for various grants, one of the questions on the application is, 'How are you sharing what you know?' This pretty much answers that question."

This article originally appeared in the November-December 2009 issue of THE Journal.