If you're a student in the US right now, chances are you've taken a lot of standardized tests.

These tests are supposed to be fair: your performance shouldn't depend on your race or gender or culture. But there are often differences between these groups' test scores, called achievement gaps.

For example, girls and boys score pretty much the same on math tests all through childhood. But once they get to be teenagers, boys start to do a little better. There are a lot of theories about where achievement gaps come from.

But there's one thing we know is going on that causes these differences in test scores, called stereotype threat. That's when test scores are influenced by stereotypes people know about their identities, even if they don't believe the stereotype is true. The easiest way to show this effect is to tell people something about the test they're about to take just before they take it— something that reminds them of groups they belong to or identify with.

For example, multiple studies have shown that if you tell groups of men and women that scores on a math test will show a gender difference before they take it, their scores will show the difference. But if you tell them that everyone does the same, the difference often goes away. You can even find this effect just by putting demographic questions about gender, race, and ethnicity at the start of a test instead of at the end.

If questions are at the start, you see a gap in scores. If they're at the end, the gap shrinks, or even disappears entirely. Researchers have found evidence of stereotype threat with pretty much any group identity associated with stereotypes.

Black students perform below white students on achievement tests if they're reminded about their race beforehand, but without the reminder, they perform the same. You can even demonstrate similar effects with older adults worried about memory loss. One study tried to see if you could push people in the other direction, too, by testing Asian-American women.

They assigned the subjects to one of three groups, then gave them a writing assignment and a math test. One group wrote about their Asian identity, another group wrote about their identity as a woman, and the third was the control group. The group that wrote about their gender had lower scores, which made sense because of the stereotype that women aren't as good at math.

But the group that wrote about their Asian identity had higher scores compared to the control group, which fit the stereotype that Asians do well in math. It's worth noting, though, that this specific study design is... kinda finicky. Two different teams tried to replicate it: one in Berkeley, California failed, but one done in schools throughout the Southern U.

S. got the same results as the original. So it might be a small or limited effect, but it also might just depend on what stereotypes are dominant in the local culture. Either way, it's clear that stereotype threat is part of why there's an achievement gap.

But why does it happen? Some research suggests that the stress and worrying about possibly conforming to a stereotype is distracting, and keeps students from being as focused as they could be. They often don't realize the source of the stress—after all, taking tests is stressful for everyone!

But even if it's unconscious, it's a stress that affects certain groups more. Other research has focused on how thinking about stereotypes might lead students to think of their abilities as fixed traits, instead of things they can work on and improve. Researchers tested this explanation by giving students a short class about how scores on standardized tests can change based on effort and motivation, compared a control that just learned about study methods.

They found that the class worked! Teaching people to think about their abilities as changeable instead of fixed encourages them to work harder to improve their math grades. But, it improved the test scores of vulnerable groups more.

Girls in the control group did worse than boys, fitting the stereotype. But girls in the class actually got better math grades than boys. Obviously, stereotype threat isn't the only thing that causes group differences on standardized tests.

There's the fact that some groups have more limited access to education, for example. Not to mention the whole complicated web of cultural weirdness that can discourage women from going into STEM fields. But stereotype threat is a separate cause that we know is there, even in students who already really like school or math!

And it won't necessarily go away just by fixing the other problems. It's also an important reminder that what you believe matters. Even though these effects were found in people who said they didn't believe the stereotypes, beliefs about their abilities were what made the difference for a lot of students.

Basically: believing in yourself is important. Science says so. Thanks for watching this episode of SciShow Psychology, and especially to our patrons on Patreon for making it happen!

If you'd like to help us keep bringing you new content like this, go to patreon.com/scishow.