TheJakartaPost

Please Update your browser

Your browser is out of date, and may not be compatible with our website. A list of the most popular web browsers can be found below.
Just click on the icons to get to the download page.

Jakarta Post

Computer-based testing can produce test bias, error variance

Students of senior high schools and vocational schools nationwide are currently participating in the annual national exam (UN)

Setiono Sugiharto (The Jakarta Post)
Jakarta
Sat, April 18, 2015

Share This Article

Change Size

Computer-based testing can produce test bias, error variance

S

tudents of senior high schools and vocational schools nationwide are currently participating in the annual national exam (UN). Equipped with advanced technological tools, some schools are poised to require their students to perform an online computer-based test (CBT). Compared with last year'€™s exam, this year has seen relatively insignificant hitches in terms of test preparation and administration.

Although disruptions have reportedly occurred in the implementation of the CBT in some schools, the exam process has in general run well, and the students seem enthusiastic about the new design of UN materials.

Admittedly, the CBT has made overall test administration become more viable and economical, and this has the potential to enhance the validity (i.e. appropriateness) and reliability (i.e. dependability) of the UN in the public'€™s eyes.

Yet, despite the progress in the UN administration this year, we cannot rule out the possibility that the CBT equally has the potential to dilute the complexities of variables other than cognitive abilities, which may impinge upon test takers'€™ performance in the test.

Unless anticipated, the implementation of the CBT faces the threat of being biased and of error variance '€” unsystematic differences among test-takers that could greatly affect their performance in the test.

Reported disruptions such as blackouts during the day of the exam, the ambiance of the rooms or other places where the tests are conducted, problems with the server, slow Internet connection and other technical glitches are sources of potential test biases.

Similarly, students'€™ familiarity with an online system as well as with working via this system affects to a greater extent their pace in doing the test.

Generally, the more familiar the students are with the system, the faster and (probably) the better their performance, compared with their counterparts who are less familiar. The familiarity with the way tests are supposed to be done in a certain test condition is known as test-wiseness.

It cannot be denied that students have different test-wiseness levels while doing the test. At this juncture, it would be reasonable to say that the cognitive factor is not the sole determinant in guaranteeing the success of doing the test.

Culture and Elementary and Secondary Education Minister Anies Baswedan has consistently articulated his stance of not using the UN scores as the sole basis for determining students'€™ passing or failure, which of course deserves to be applauded.

Other scores accumulated from students'€™ quizzes, mid-term and final school exams will also be taken into account. These scores, together with the UN score, will constitute the final score.

However, in spite of this new policy treating the UN score as part of other scores, there is no reason for not taking students'€™ UN scores seriously and for not finding out factors affecting how such scores are obtained.

While attention has so far been devoted exclusively to the seeking of right formats and designs of the exam, little has been done regarding whether or not the scores obtained are the results of a valid interpretation of students'€™ answers.

Furthermore, social implications brought about by the exam have never been touched upon and reveled to the public at large.

The bureaucratic silence over these implications gives the impression that the UN is a sociopolitically free educational instrument intended simply for measuring learning outcomes. Its political motives are kept hidden.

It is important to note here that the UN, as a standardized test, is not constructed in a vacuum and devoid of sociopolitical considerations. The fact that its implementation is always sustained, despite endless controversies over it, clearly reflects a sociopolitical side of the exam.

Anies'€™ school-sensitive policy seems to have earned respect and trust from education practitioners, teachers and parents, as evidenced by the dwindling oppositional voices from the public and from the relative smoothness of the exam processes.

Yet, the overall final results of the CBT, its social implications, usefulness and its effects on learning and teaching have yet to be seen, before any claim of its success can be made.

________________

The writer is an associate professor of English at the school of education and language, Atma Jaya Catholic University, Jakarta.

Your Opinion Matters

Share your experiences, suggestions, and any issues you've encountered on The Jakarta Post. We're here to listen.

Enter at least 30 characters
0 / 30

Thank You

Thank you for sharing your thoughts. We appreciate your feedback.