This is a re-post of Dennis Christopher’s popular “Item Ordering on Tests to Reduce Stress” originally posted July 2, 2015
Your child went to class, completed homework, and studied.
He or she arrived at the exam confident about the material.
But if he or she has test anxiety, a type of performance anxiety,
taking the test is the most difficult part of the equation.
“Test Anxiety | Anxiety and Depression Association of America, ADAA,” n.d.
As an assessment specialist I know that best practices in testing is to randomize test question delivery by content, or student learning expectation (SLE). By randomizing the delivery there is an added level of security and less of a possibility that one answer to a question will cue the answer to the next question.
And as a life-long educator I can understand the reasoning. We need to find out what the student knows and is able to do without giving away the answer. To achieve that goal, random ordering makes sense.
However, we do not teach randomly and we do not think randomly. Is there a case to be made that some students would perform better on tests if we did not impose random order on test items? That’s a topic I explore in greater detail in this white paper. For the purposes of this blog post, I will focus on how the order of test items might reduce test-taking stress.
When we teach students, we do not jump from the study of the stars one day to the circulatory system the next day. That would be random. As we talk to friends and colleagues, we do not talk about what our kids did today in one breath and our clogged garbage disposal in the next. (Well, I guess we would if our kids caused the clogged the disposal with one of their toys.) In most instances, though, this also would be random.
For every topic we discuss, our brain has to pull together all the information it knows about that topic so that we can have a coherent conversation. Whenever the topic changes, our brain has to switch gears and pull information related to the new topic. These groups of connected information are called mental sets. Every time a topic changes, we have to pause to pull a new mental set together. The more quickly the topic is changed, the more taxing it is for the brain to keep pace.
Now let’s look at a test again. We deliberately build our tests with randomly ordered questions. And every time a student brings up a new question he must take the time to pull together a new mental set. The more time it takes to gather that information, and the more unrelated each item appears to be, the greater the stress on the student.
On the other hand, a study conducted by Pettijohn and Sacco (2007) found that grouping items by content and sequentially ordering them made a difference for a select group of test takers identified as “students who experience intense test anxiety.” When these students were asked about their level of test anxiety when given sequential (S) and random order (RA) item exams, their answers pointed the researchers to the following conclusion:
“Our post-exam anxiety assessments showed that participants reported greater anxiety in response to the RA (random order) test question order condition compared to the S (sequential) condition…. Overly anxious students, students with learning disabilities, or students with other special circumstances may benefit from sequential test question order.”
Students who take sequentially ordered exams appear to experience reduced levels of stress. As discussed in a white paper produced by my Questar Assessment colleague, Dr. Les Sewall, and I titled “Using Performance Assessments to Bridge the Gap between Accountability and Learning,” critics of standardized testing often express concern about the stress these exams place on students. That white paper, along with my other white paper titled “The Effects of Item Order on Student Scores,” outlines ways to reduce students’ stress before and during the exam. Sequentially ordered items on exams, along with teaching students memory, retention, and retrieval skills, may reduce the stress on the student and increase test scores.
Although states and school districts may emphasize random order testing for security purposes, we need to also consider the effect of random order testing on students’ stress levels and scores. Would it make sense to allow alternative students and students with high test anxiety to take a test with sequentially ordered or content grouped items?
Pettijohn, T. F., & Sacco, M. F. (2007). Multiple-choice exam question order influences on student performance, completion time, and perceptions. Journal of Instructional Psychology, 34(3). Retrieved from http://www.freepatentsonline.com/article/Journal-Instructional-Psychology/170156969.html
Test Anxiety | Anxiety and Depression Association of America, ADAA. (n.d.). Retrieved from http://www.adaa.org/living-with-anxiety/children/test-anxiety