Can We Teach Kids to Be Kinder?

Kindness is typically understood as "actions intended to benefit others, usually at some cost" to the actor. Research has found that kindness-related traits emerge early in infancy and continue to develop throughout childhood. Individual differences in kindness—and related traits such as altruism, benevolence, and compassion—are partly due to genetic differences, but largely due to differences in children's experiences outside the home.

School-Based Social and Emotional Learning

Schools are one possible source of these experiences, and school-based social and emotional learning (SEL) programs are one attempt to provide them. A growing body of evidence suggests that they work. Recent meta-analyses have found that SEL programs have small-to-moderate positive effects on prosocial behaviour, as well as on well-being, attention, behaviour, peer relationships, and academic performance.

We recently conducted a randomized controlled trial of our own SEL program, Learn Kind, which is currently taught by more than 1,000 teachers to more than 100,000 students each year.

Learn Kind is an eight-part kindness-based curriculum for students in grades K–8. Rather than simply telling children to "be nice," the program treats kindness as an object of scientific inquiry. Students are invited to act as "student scientists," discovering for themselves what increases kindness and the effects it can have. The curriculum covers topics including the science of kindness, compassion, cooperation, self-kindness, and the relationship between helping others and happiness. It also includes an experiential component in which students perform acts of kindness and record their effects on themselves and others, testing the hypothesis that helping others makes people happier.

To evaluate the program, 50 teachers, reporting on more than 1,000 students, were randomly assigned either to teach Learn Kind during the autumn term (the treatment group) or to wait until the spring term (the waitlist control group). Teachers rated their students' kindness (using a simple kindness grade of A, B, C, D, or F), prosociality, emotions, attention, behaviour, peer relationships, happiness, and academic performance in mathematics and reading.

Study Findings

The study found significant improvements across a wide range of outcomes. Most strikingly, students in the Learn Kind group improved by approximately one full letter grade in kindness, moving from an average rating of C to B. Compared with the control group, they were also rated as happier, more prosocial, better behaved, and more attentive; had stronger peer relationships; and showed modest improvements in mathematics and reading. Across all outcomes, the average effect size was d = 0.34.

Importantly, these improvements largely persisted. When students were assessed again three months later, gains in kindness, happiness, behaviour, peer relationships, and academic performance were still evident.

Benchmarking these results against previous evaluations suggests that Learn Kind ranks in approximately the top quartile of SEL programs.

These findings add to growing evidence that kindness is not simply an inborn disposition but a skill that can be developed through experience. More broadly, they suggest that well-designed educational programs can improve not only kindness itself but also children's social, emotional, and academic development.

We hope this is only the beginning. Our plan is to expand Learn Kind by incorporating classroom experiments exploring additional virtues, such as gratitude, generosity, fairness, and honesty. This would create a "virtual laboratory" in which thousands of student scientists contribute to the psychological science of character development, with the potential to identify ever more effective interventions that help kids become kinder and happier.