Incentivizing Self-Skepticism On School Tests

Student answers to exam questions usually have two possibilities: they are either right or wrong. Sometimes the question is worth multiple points and it is possible to get “part marks” by getting some things right and some things wrong. For the purposes of this post, I’ll ignore situations where it is possible to get part marks. Let’s just assume questions are the fill-in-the-blank or true-or-false type each worth 1 point.

What if school tests offered a third option?

You get 1 point for getting the answer right. You get no points for leaving the answer blank. You lose 1 point for getting the answer wrong.

This gives students a reason to “think before they speak” – to distinguish between answers they’re more than 50% sure of and those that they’re less than 50% sure of. Rather than taking a wild guess, students will learn the skill of assessing the likelihood of their guesses being correct and holding off if they aren’t certain enough. Each question is a reminder to think hard about whether they've really got the right answer. It's a call for self-skepticism.

I’m not sure whether this would have good consequences or not. It could train people to be sheepish, risk-averse, and obedient. I’m also not sure the 1-0-1 point system is optimal. Maybe it should be more like 2-0-1 so that guesses won’t be too harshly penalized. I think incorporating incorrect guesses into scoring systems could potentially have good effects on students.