The social psychologist Joshua Aronson and I did one of the early experiments testing this. We invited white and Black Stanford University students into our laboratory one at a time and gave them a very difficult thirty-minute section of the verbal Graduate Record Examinations (GRE).

In a third condition, the feedback-giver said to each participant that he was using high standards in evaluating their work and that, having looked at their essay, he believed that they had the potential to meet those standards. This signalled to the Black participants that they weren’t being stereotyped. If they were, they wouldn’t be held to high standards or have their potential affirmed. This feedback was water on parched land for Black participants. They trusted it as much as white students did. And trusting it, they used it to improve their essays. Three times more of them took their essays home to improve than did participants in any other condition of the experiment.

They were left with that nagging ambiguity: Was it my essay or a bias of the feedback-giver that shaped the feedback? In these two conditions, Black students wanted to trust the feedback, be motivated by it. But there was that ambiguity. They remained in churn, protectively less engaged. Take this ambiguity away— as did the wise feedback-giver— and Black students showed strong trust and motivation. Their capacity to trust and be motivated was there all along. The spark they needed was evidence that the feedback-giver’s expectation of them was not limited by the stereotype about their group’s abilities.

Gregory Walton and Geoffrey Cohen tested the effect of a similar intervention at the college level. They exposed freshmen to a persuasive message that presented the social stresses of college life as transient and as something everyone experiences. This message improved Black students’ grades over their next three years in college when compared to Black students who didn’t get this message. How did it do this? Further analysis showed that it helped Black students see college stress as something not peculiar to their identity. It reduced their identity threat. They could trust their environment more— churn less— and their grades reflected it.

For students facing Tyler-type identity threat, they improve performance less by directly affecting their learning than by reducing the identity threat that can otherwise interfere with their learning. That’s their route of action, their shared active ingredient. They do it in different ways. But they all change the socio-psychological experience of school for these students. And throughout the now extensive literature testing these effects, it is for these students that these interventions have their largest, most reliable effects.