The increasing use of computers in schools and classrooms has prompted educators, testing experts and test developers to look at ways of applying technology to student assessment. Many supporters argue that computerized exams are more in line with current instructional practices and the way students are accustomed to learning. Interest also has been piqued as states place greater emphasis on assessments, the results of which might determine grade promotion or high school graduation for students, and rewards or sanctions for schools. While pencil-and-paper tests remain the standard form of test taking, the hope is that technology can provide a more efficient and possibly more effective way to administer student exams.
While computer-assisted testing at the school and district level has been around for several years, broad-scale use is still in its beginning stages. Testing companies have used computers to administer professional exams and, increasingly, graduate-level exams are being taken via computers. A few states, including Oregon and Virginia, are moving toward — or at least piloting — online state assessments (see What States Are Doing on this issue site). Other states are considering a similar direction.
A limited number of studies have examined whether test results differ by the mode of administration — written vs. computer. While the studies did not indicate test result differences for multiple-choice questions, they showed a mixed picture for open-ended test questions - from no effects to large effects in student performance between the two modes. Computerized math exams seem to pose the most trouble for students. In general, students who can keyboard at a moderate to fast pace, and are accustomed to writing on computers, tend to do better on computerized tests than on paper exams. Conversely, students who keyboard at a slower pace and don't use computers for writing assignments tend to do better on paper exams.
To supporters, these results reflect the gap between the growing use of computers in schools and the testing strategies used for evaluating performance, and thus supporters see great potential for computerized exams. To those who are more skeptical, the results highlight the "digital divide" between students who have access to computers and those whose access is limited.
Using computers to administer exams has several advantages, including: immediate reporting of scores, electronic transfer of test results and the ability to tailor questions and content. Currently, teachers often struggle to receive test results in a timely manner that allows them to adapt teaching methods and curricula for individual student needs. The quick turnaround for test results could address the timeliness problem, as well as provide more flexibility to administer and score exams on an as-needed basis.
The primary disadvantage is the cost associated with start-up and maintenance of technology and the database of test questions. In addition, there are concerns about losing assessment data if computer systems crash, providing adequate technical support at schools and delivering high-quality, affordable exams. Equity might become an issue if students with more access to computers and who are skilled keyboarders have a greater advantage with computer-based tests than those who lack access and skills.
Several types of computer-based tests are available, and testing experts often make a distinction depending upon the extent to which the exam is adaptive to a student's performance. According to the College Board, the degree of adaptivity ranges from linear to adaptive. With linear tests, the rigor of questions do not change based on the student's performance, while adaptive tests are exams in which each item presented is dependent on the student's performance. For example, if a student answers a question correctly, the computer selects a more difficult question from a database, and then continues to alter the rigor of questions. Three other types of tests fall in between these two on the continuum.
Despite concerns, the use of computers to assess students seems to be gaining attention and most likely will become a more common form of test delivery. Educators, policymakers, assessment experts and testing companies must work closely to minimize the disadvantages and maximize the advantages if computerized tests are to become an efficient, effective and equitable way to assess student performance.