For most students, science, math, engineering, and technology (STEM) subjects are not intuitive or easy. Learning in general—and STEM in particular—requires repeated trial and error, and a student’s lack of confidence can sometimes stand in her own way. And although teachers and parents may think they are doing otherwise, these adults inadvertently help kids make up their minds early on that they're not natural scientists or “math people,” which leads them to pursue other subjects instead.

So what's the best way to help kids feel confident enough to stay the STEM course? To answer this question, I spoke with Carol Dweck, a professor of psychology at Stanford University in California. Over the past 20 years, Dweck has conducted dozens of studies about praise’s impact on students’ self-esteem and academic achievement. Here is a transcript of our conversation, which has been condensed and lightly edited.

Alexandra Ossola: What sparked your interest in this field?

Carol Dweck: We undertook this research at the height of the self-esteem movement, when the gurus were telling parents and teachers to praise kids to the hills, to tell them how talented, intelligent, brilliant they were. And this was supposed to boost their confidence and set them up for a successful life. And I thought, I don’t think so. When students thought of their intelligence as a thing that’s just fixed, they were vulnerable. They were not willing to take on challenges that might test their intelligence, and they weren’t resilient when they came into obstacles. Would children want to take on challenging things if they think that’s their claim to fame? It would discredit this valuable, permanent quality.

Ossola: What is a mistake that parents and teachers often make when it comes to praising kids?

Dweck: They often praise the ability, the talent, or the intelligence too much. The opposite of this is the good process praise. This is praise for the process the child engages in—their hard work, trying many strategies, their focus, their perseverance, their use of errors to learn, their improvement.

We conducted a study where we recorded videotapes of mothers interacting with babies when they were one, two, and three years old. The more the mothers gave process praise, the more their kids had a growth mindset and a desire for challenge five years later. And now we’re finding how much better those kids are achieving even two years after that.

Gender is relevant: The mothers are giving boys more process praise than they’re giving girls. Years ago we found the same thing in teachers giving feedback.

It’s not that they’re necessarily favoring boys. Even as infants, girls are more socially tuned in, easier to engage, and more able to maintain an interaction. This means that parents may have to work harder to engage boys and keep them engaged in a mutual activity. Parents’ harder work could take the form of more process praise—more narrating and commenting on what the boy is doing, more praise for the strategies they boy is trying, and more praise for attention and persistence.

It’s easy to get a girl to focus on you, and adults react to that. I’m not saying there aren’t stereotypes of girls that aren’t coming in. But this difference could really begin in a seemingly irrelevant biological difference between boys and girls that draws more process focus in parents and teachers, plus teaches boys a good lesson: when you pay attention and work hard, you do better. With girls, who are just self regulating away, they don’t need to give us those lessons. But as a result they think that when adults give them criticism, it’s for something really important. It’s about their ability, how good they are at something.

Ossola: What is the right balance of praise? How much is too much?

Dweck: That hasn’t been as well researched as what to praise. But we did some work, which we haven’t published, and we found that when we gave kids lots and lots of praise then discontinued it, they either lost motivation or they did a variety of strange and distorted things to get the adults’ approval back. There is such a thing as too much praise, we believe. When you praise someone, you are making their actions and performance yours. So they’re looking over their shoulder and not owning their work.

Ossola: Does praise affect people at different places in their academic lives? Can it carry on into adulthood?

Dweck: It can be powerful at any age. Being a good parent has become synonymous with giving out ability praise. Parents still think this is the greatest gift they can give to their children, and as a child gets more and more insecure, they give more and more of it. And, by the way, a lot of employers and coaches have said, “My employees cannot get through the day without accolades and validation.” Even professional coaches have said they cannot give feedback without these people feeling that they’ve crushed them. We’ve created several generations now of very fragile individuals because they’ve been praised and hyped. And feel that anything but praise is devastating.

Ossola: How can parents and educators foster this resilience to obstacles in other ways?

Dweck: Actually, praise may not be the optimal way, but we are so praise oriented. We can ask the child questions about the process: “How did you do that? Tell me about it.” As they talk about the process and the strategies they tried, we can appreciate it. We can be interested in it. We can encourage it. It doesn’t have to be outright praise.