Get important education news and analysis delivered straight to your inbox
When you’re running out of time on a multiple-choice test, it makes sense to guess the rest of the answers rather than leave the questions blank. But it turns out that a surprisingly high number of students are guessing their way throughout a test, even when they’re not pressed for time or trying to boost their scores.
Testing experts have a name for it: rapid guessing. One researcher at NWEA, an organization that produces standardized tests, says he has come up with a way to detect rapid guessing and has found that it is particularly prevalent on reading tests. Boys, in general, do it at twice the rate of girls. By eighth grade, nearly 16 percent of boys are guessing their way through at least 10 percent of a commonly used reading assessment in U.S. schools.
“If you’re disengaged during a test, one of the the things you may do, particularly if it’s a multiple choice test, is start entering answers rapidly, just to get this thing over with because you don’t want to do it,” said Steven Wise, a senior research fellow at NWEA.
The consequence is that many students are posting considerably lower scores than they otherwise would be if they were paying attention and putting in full effort, according to Wise’s calculations. Students need to cooperate and try their best in order for testing companies to produce accurate scores, Wise explained.
“In a way, we’re putting ourselves out on a limb. We’re saying all of our scores aren’t trustworthy,” said Wise, who was previously a professor of assessment and measurement at James Madison University. “It’s not just us. Disengagement is going on with all assessments.”
For teachers and schools, rapid guessing makes it harder to know what students are really learning in school. And it calls into question all kinds of achievement gaps, including differences in test scores between boys and girls and between black and white students.
Thanks to computerized testing, Wise believes he can detect whether a student is engaged based on the number of seconds it takes to answer a question. If a student answers a question in less than 10 percent of the time that it took an average student to answer a question, he classifies it as a “rapid guess.” Depending on the complexity of the test question, that rapid-guess threshold can be less than two seconds or as long as 10 seconds. (Wise said this was a conservative threshold that would not accidentally lump exceptionally rapid thinkers in with the rapid guessers.)
Wise analyzed the test scores of almost nine million students in grades two through eight across the country who took NWEA’s Measures of Academic Progress (MAP) test. In a June 2019 report, Wise wrote that rapid guessing was higher for test questions that contain more reading or that are perceived as more taxing. For male test takers, rapid guessing increases with each grade level. (See the accompanying bar charts for guessing rates by gender and subject.)
Black students tend to make rapid guesses 50 percent more often than white students do, Wise said in an interview. That phenomenon has led Wise’s colleague, James Soland, to question how much of the black-white achievement gap is actually a gap in how much effort students of different races are putting into the test. In a 2018 analysis, Soland said that it was hard to quantify exactly how much of the achievement gap between black and white students is an effort gap and he encouraged more research into why effort on a test differs by race.
One might wonder if kids are guessing when they really don’t know the answer to a problem. But Wise’s research points to student motivation as the cause. That’s because he looked closely at 30,000 students who happened to take the same standardized test two days in a row. For whatever reason, the student wasn’t in the mood on the first day and rapidly guessed through a large portion of the test. On the retest, students typically took more time with each question and posted a “markedly higher” score. It’s unlikely that a student would suddenly know a lot more the next day and that’s a sign that the first day’s guessing was more psychological than academic.
Why boys are more prone to rapid guessing than girls is a matter of conjecture. The field of psychology has developed several theories. Some observe that girls tend to be more “conscientious” and “agreeable” than boys, and these psychological traits are (negatively) correlated with rapid guessing rates. Another theory focuses on work avoidance, which boys are more prone to admit to on psychological surveys. “Boys may look at a test item and say, ‘I don’t feel like doing that,’ ” said Wise.
Identifying rapid guessing has become a new business line for NWEA. The organization recently began selling a service that can alert a teacher on his computer monitor in real time if a student starts to engage in rapid guessing. There isn’t a formula for getting a student to focus again, but Wise said that teachers have been effective in curbing rapid guessing during a test. “We left it to the proctor to use their discretion for the best way to help re-engage Johnny because they know Johnny, and what works and doesn’t work with with him,” said Wise.
Wise has more concrete advice for teachers and school leaders before a test starts. He suggests that teachers be more aware of testing conditions, especially noisy distractions, and avoid scheduling tests just prior to lunch.
Beyond logistics, the techniques of social-emotional learning might be helpful in boosting motivation. Wise believes it’s helpful to ask students to set “growth goals” at the beginning of the year so that they take an interest in seeing their test scores improve over time.
“I’m not naive to think this always works,” said Wise. “But giving them ownership of the process is one way of trying to get students engaged.”
This story about computerized testing was written by Jill Barshay and produced by The Hechinger Report, a nonprofit, independent news organization focused on inequality and innovation in education. Sign up for the Hechinger newsletter.