Because of increasingly affordable technology and wider access to high-speed broadband Internet connections, more companies in the computer-based-assessment industry are devising ways to deliver tests online. But many obstacles still hamper the use of those assessments for large-scale testing.
“There’s no doubt ... that online testing is really the most productive and reliable way [to administer assessments],” says Daniel Wakeman, a vice president and the chief information officer of the Educational Testing Service, the Princeton, N.J.-based nonprofit assessment company that administers the SAT and other major tests.
Improved “caching” technology, which retains information even if the Internet connection drops and sends it when the connection is re-established, as well as improved Internet stability and increased bandwidth capabilities, helped the ETS administer its Test of English as a Foreign Language, or TOEFL, online.
In addition to reaching more people, administering the test through the Internet has cut back on the subjectivity of the test, says Tom Ewing, a spokesman for the ETS. For example, during the speaking portion of the TOEFL, students could potentially garner clues about their performance from the reaction of the evaluator, something that doesn’t happen when the test is taken on a computer, which exchanges computerized audio files through the Internet for this portion of the test.
“It’s a much more valid indicator of the test-taker’s ability than the face-to-face exam,” Ewing says.
Some of the challenges that the ETS faced when implementing the online version of the TOEFL parallel obstacles that school districts face when attempting to conduct computer-based testing on a large scale, Wakeman says.
“Many schools lack the proper infrastructure to support wide-scale computer-based testing,” he says. They might not have enough computers or bandwidth to support an online assessment, he says, and upgrading the system to support those applications could be expensive.
The Upper Saddle River, N.J.-based Pearson Education, a leading educational publishing company, released an online version of its Stanford 10 assessment, a multiple-choice exam for students in grades 3-12, in October.
The assessment provides immediate feedback for teachers, allowing them to use the results to adjust instruction, says Deloris Flint, the director of products for the company.
In addition to trimming shipping and printing costs and eliminating the wait between when students take the test and when they see the results, the assessment helps prepare students for the future, says Peter Frey, the technology-strategy director for the educational assessment group of Pearson.
“We know that as children get prepared for careers and industry that everything is going to be computer-based,” he says. “We believe that at a fundamental level, online assessments really enable students to better prepare for the future.”
Computer-based delivery also changes the relationship between assessment and instruction, says Frey.
“We typically separate out instruction and assessment,” he says, but with computer-based assessments, “we’re blurring the lines.”
But not all companies tie the technology to administer computer-based tests with the content of the assessment.
In the case of Virginia’s 168,000-student Fairfax County public schools, school officials have formed a partnership with Northrop Grumman, a Los Angeles-based defense and technology company, to provide the technological framework for computerized tests.
"[Fairfax County schools] wanted a tool that would allow teachers to work together to develop their own assessment items for their classroom formative assessments and use that data to identify where students are struggling and to refine instruction in the classroom,” says Joseph W. Warden, the director of business development for the company.
Allowing schools and teachers to add their own assessment items to the technological framework—in this case, called ASPIRE—gives schools more localized control over the assessments and a greater understanding of what the assessment is measuring, says Warden.
“A lot of our competition are publishers, and their main goal is to develop items or questions—and they’ve been writing great questions—but our feeling on that is that they may be good for benchmark assessments, but that’s not true for formative assessments,” he says.
A version of this article appeared in the January 21, 2009 edition of Digital Directions as Assessment Connection Inc.