As long as someone is already perceived as competent, acknowledging mistakes can be beneficial.

For another set of students, the researchers worded the praise slightly differently. Rather than praising the person, or telling them how smart they were, the researchers praised the process, or how hard they were working (“You must have worked hard at these problems”).

Telling someone they’re smart, good at math, or a great presenter implies that their performance depends on a stable trait. If they did well on a test, they have that trait, but if they did badly, well, they’re just out of luck. They don’t have what it takes and there’s not much they can do to change it.

But rephrasing that feedback as process praise is more likely to have the intended effect. Telling someone they did well, or did a good job on a test or presentation, focuses less on stable traits and more on the particular instance at hand. Which means if things don’t go so well once in a while, it’s not a mark of failure or lack of ability. It’s just a misstep and a reminder to work harder next time around.

Academic studies point out that there is a “compensation effect” between warmth and competence: people tend to assume that people who are warm are incompetent and those who are cold, competent.

I’d go so far as to say that insisting on high standards without psychological safety is a recipe for failure—and not the good kind. People are more likely to mess up (even for things they know how to do well) when they’re stressed. Similarly, when you have a question about how to do something but don’t feel able to ask someone, you’re at risk of running headlong into a basic failure. Also, when people encounter intelligent failures, they need to feel safe enough to tell other people about them. These useful failures are no longer “intelligent” when they happen a second time.