Online Assessment

Uncharted Territory

• By Matt Villano
• 03/01/08

Are you choosing the right online assessment products and getting the most out of the tools you have? Online assessment is fraught with pitfalls, but these savvy educators and technologists are meeting the challenge-and then some.

There's certainly no shortage of online learning platforms out there today. Blackboard, Desire2Learn, Sakai, Moodle, Angel Learning, and Datatel (projected to be a CMS player in Q4 2008): You name the interface, and chances are that someone at your school has evaluated it at some point in the not-too-distant past. But investigating the value of the assessment components of these tools, now that's another story altogether. This exploration-essentially, the process of assessing online assessment-can be far more complicated. And while many higher education administrators trust their CMS vendors implicitly, a growing number are engaging in their own forms of metrics to gauge how well students are doing when they're educated or accessing education content online.

Some officials see this process as a critical part of online learning systems. Others see it as an act of calculating return on investment (ROI)-a way to see precisely how much bang they are getting for their buck. Ron Legon, executive director of Quality Matters, a program designed to certify the quality of online courses and online components, says that no matter how educators assess their online assessment tools, it's a critical part of performance evaluation overall (see "Setting the Standards"). (Quality Matters is run by MarylandOnline, a consortium that champions distance learning in Maryland and serves as a directory for Maryland schools involved in the online learning experience.)

"To offer online learning is one thing," says Legon. "To actively evaluate it to make sure it's doing its job, is something entirely different."

Selecting Rubrics and Metrics

Inherently, assessment tools or rubrics are nothing without metrics. In traditional classroom settings, most of these metrics take the form of test scores, compiled after a particular lesson (in the case of formative assessments) or a particular sequence of the curriculum (in the case of summative assessments). Other assessments consist of grades or rankings for things such as participation, homework, and attendance. This is nothing new.

Many of these same metrics exist in the world of online tools, as well-the media by which they are applied are just different. Instead of distributing a paper exam, for instance, a professor may have students respond to multiple-choice questions via a web browser. Instead of having students meet at the library for group homework assignments, a professor may require them to meet in an online collaboration environment.

Most professors apply metrics through predetermined assessment rubrics. At Rio Salado College (AZ), however, many of the rubrics are fun: multiple-choice practice quizzes turned into the form of online games with a little help from Quia web-based software. Jennifer Freed, Rio Salado faculty chair of instructional design, says the playful interface gives students a chance to learn comfortably.

"The games are fun and they provide instant feedback," says Freed, who notes these formative assessments are interspersed with more "serious" webbased summative assessments once or twice throughout the semester. "I can't think of a better way for students to process new material."

At Newbury College (MA), "metrics" are much more conceptual. Yes, educators assign scores to certain tests and assignments, but at least in certain psychology classes, Professor Charlie Virga is more interested in seeing that his first- and second-year undergraduate students can demonstrate the "construction of knowledge" from the beginning of a semester, to the end of it.

For Virga, this means careful scrutiny of online discussion posts. With the help of his school's Blackboard system, he archives every post and grades them periodically throughout the semester. Relevant posts that link to course material and provide elaboration or additional information receive the highest marks. Irrelevant posts, and posts that have no link to course material or personal experience, receive no score.

"In my book, it's all about critical thinking," he says of his rudimentary rubrics. "I don't have access to [my students'] thought processes online, but by looking at the discussion posts, I can try to identify the turning point where they started to see something that they couldn't see before."

Keeping Tabs

With a course management system such as Newbury's, archiving data on performance is a cinch. Such is the case with many other CMS platforms and online assessment tools, too. Collecting data on student performance in the virtual environment, however, is only half of the assessment effort; once professors have the data, the next key step becomes figuring out how to make sense of it.

One way to keep tabs on the degree to which students are interacting with online assessment technologies (and with peers via the tools) is to apply business intelligence. With the help of a virtual learning environment from L Point Solutions called Inetoo, professors can encourage student collaboration and communication online, and later log in to analyze how students interact with content and with each other.

This service, dubbed "performance intelligence," is something that founders Robert Brouwer and Ahmed Abdulwahab say is a higher education spin on the kind of business intelligence used by companies in industrial and manufacturing sectors. While this product is brand-spanking-new, Paul Kim, a professor at Stanford University (CA), is wasting no time deploying it; he's planning to pilot it in his Web-Based Technologies in Teaching and Learning class this spring.

"After the completion of this course, students will be able to describe how web-based communication, collaboration, and visualization technologies play a role in the behavioral, cognitive, constructivist, and social dimensions of learning," says Kim, who also serves as CTO of the university's School of Education.

Finally, at the Rose-Hulman Institute of Technology (IN), educators have turned to the Learning Management Suite from Angel Learning to map various content items (such as assessments, drop boxes, and discussion forums) to institution-wide and coursespecific objectives and to generate reports based on student performance related to all items associated with a given standard or objective.

Claude Anderson, professor of computer science and software engineering, says the school recently has incorporated Subversion for storing and communicating all of its faculty-level course assessment documents, and for charting version control.

"We used a wiki-based system for a couple of years, but found it too cumbersome," says Anderson. He adds that with Subversion, Rose-Hulman has "greatly simplified the coordination between various faculty members teaching a course."

Dissuading Cheaters

In a brick-and-mortar classroom, it's easy for teachers to catch students peering down at a cheat-sheet or passing answers to a pal. In a virtual classroom, however- where in most cases educators have never seen students face-to-face and have no idea what kinds of technology setups students have in their homes-sniffing out cheaters is a much more difficult task.

This is a challenge Karen Swan knows all too well. As research professor for the Research Center for Educational Technology at Kent State University (OH), Swan works regularly with professors to devise ways to prevent cheating in the online world. Yet, the harder she tries, she admits, the harder she finds the task. Her solution: keeping students active with assessments before, during, and after every class.

Extreme? Perhaps. But as Swan sees it (after years of research), short of locking students into a particular browser (which still isn't foolproof if students have a second computer at home), there is no way to tell if online students are working together behind the scenes. Rather than trying to prevent this, she argues it's better to throw multiple and repeated assessments at students so-at least at some point- they are forced to do their own work.

"The only feedback for whether or not they're learning is the assignments they do, and because you don't have people nodding their heads in a classroom [as you teach], those [assessments] should be multiple," she says. As for assessing the quality of the feedback, Swan concedes it's not her priority. "As long as I'm getting feedback, I'm happy."

Other educators agree. Virga, the psychology professor at Newbury, says that in most online classrooms, since it's so difficult to catch cheaters in the act, educators simply must assign assessments and trust that students won't cheat. He adds that by not having a physical classroom to which students must report, educators can get away with requiring additional assignments, thereby getting a better sense of who and what each student is all about.

"In a face-to-face class, all you're actually getting is their papers," quips Virga. "In an online class, it's paradoxical, because even though they're not there, you can demand and expect more."

Improving Assessment

Looking forward, perhaps the best way to assess the performance of online assessment tools over time is to embrace evolution. The easiest way to do this is simply to stay on top of recent research into online assessments, a chore that is perhaps best accomplished by keeping abreast of the latest publications that deal with the subject (see "For More Information").

On individual campuses, there are other, more proactive options for implementing the latest and greatest in online assessments. Some educators, such as those at Rose-Hulman, administer surveys to all students who participate in online learning, and go through survey responses at the end of every semester to see how they can improve the online assessments and the web-based learning experience overall.

Educators at Rio Salado are even more meticulous: At the end of every school year, Freed says instructors look back at each individual assessment and compare student performance on every question. If a majority of students got a question wrong, educators may go back and tweak the wording or rewrite the question altogether. If a majority of students got a question right, educators might make the query more challenging.

"More than anything, we want to make sure that assessments align with what we're teaching," she says, noting that the process is indeed time-consuming, frequently daunting, but still worth it because of its impact on the quality of the education delivered. "In the end, the curriculum is more important than [the work on] any assessment or online interface."

::WEBEXTRAS ::
Electronic Student Assessment: The Power of the Portfolio. Case Study: Seton Hall (NJ) Embraces Assessment with Technology.

Matt Villano, senior contributing editor of this publication, also writes regularly for The New York Times, San Francisco Chronicle and Sunset. He is based in Healdsburg, CA.