While it can be better than facing a totally blank sheet of paper, the multiple-choice quiz can still provoke a lot of anxiety and uncertainty. Students look at their options with doubt, wondering if their first instinct is the correct one or if they should double-back to fill in another circle.

Naturally, knowing the correct response is best. But if you’re in a situation where you have to guess, there’s a good way to improve your odds of getting it right. Writing for Quartz, Justin Couchman, Ph.D., an associate professor of psychology at Albright College, says that our first instinct is not as reliable as we think: Something called “endowment bias,” which is a preference for the first thing to pop into our heads, can make us too attached to our impulsive responses.

To find out whether a student’s instinctual answers stood a better chance of being correct, Couchman had his class take a psychology exam—for a real grade—and mark each answer to indicate whether they were confident in the choice or just guessing. They also indicated whether the original response was altered. Couchman found that revising the first answer gave students a better chance of being correct than if they stuck with their gut-instinct first response.

In another experiment, Couchman had students rate confidence in their answers on a scale from one to five but had them stick with their original answers—no revisions were allowed. The ones students were most confident in tended to be correct.

The takeaway, according to Couchman, is that students were able to assess their feelings on a given question in the moment and then return to them for possible revision. By looking at their own self-assessment—a form of metacognition, the act of thinking about thinking—they could choose to alter answers they had the least amount of confidence in. It’s a more objective way of gauging answers than following a blanket rule of your first instinct usually being right.

The next time you’re faced with several possible answers, rate your confidence level for each. If it’s high, stick with your initial gut response. If it’s low, revise. You’ll stand a much better chance of success than if you followed a blanket first-instinct rule or an always-revise rule.

[h/t Quartz]