Do affirmations actually work? What the research says

What is self-affirmation theory?

The research most people mean when they ask whether affirmations work grew out of self-affirmation theory. Claude Steele introduced it in 1988, proposing that people are motivated to maintain a sense of themselves as adequate, competent, and able to handle what matters. When one part of that self-image is threatened, affirming an unrelated value you care about can restore the whole, which reduces defensiveness rather than resolving the threat head on.

A 2014 review in the Annual Review of Psychology, by Geoffrey Cohen and David Sherman, synthesized decades of this work. Their conclusion is careful: brief exercises, usually writing about a core personal value, can lower defenses and let a threat loom smaller, and well-timed ones have improved outcomes in education, health, and relationships, sometimes for months. They are explicit that these are not magic bullets. Effects depend on timing, context, and whether the moment is one where a threat is already undermining the person.

What does the research actually show?

The strongest evidence is for affirmations that ease a specific pressure. In a 2013 randomized experiment published in PLOS ONE, Creswell and colleagues found that writing about an important personal value improved problem-solving under time pressure, but only for people who were high in chronic stress and would otherwise have underperformed. It did not boost people who were not stressed. The benefit is buffering, not a general lift.

A 2006 field study in Science, by Cohen and colleagues, ran a brief values-affirmation writing exercise with 7th-graders. It raised the grades of African American students and narrowed the racial achievement gap by about 40 percent, with the largest gains for lower-performing students and no meaningful effect for European American students. A 2016 neuroscience study in Social Cognitive and Affective Neuroscience, led by Cascio, used fMRI to show that reflecting on future-oriented core values activated the brain's self-processing and reward regions, a mechanism finding rather than proof of broad life change.

Pooled across many studies, the effects are real but small. A 2023 meta-analysis in Healthcare of 144 education studies found a modest average effect and a lot of variation, stronger for identity-threatened groups and in-person delivery, weaker or negligible otherwise. A 2015 meta-analysis in Health Psychology, by Epton and colleagues, found small but reliable effects on health-related intentions and behavior, working by reducing defensive resistance to threatening information. Large replications add caution: Hanselman and colleagues in 2017, and Flynn and Bordieri in 2020, both failed to reproduce classic effects, which appear fragile and dependent on conditions we do not fully understand.

When do affirmations backfire?

This is the part that gets left out. In a 2009 study in Psychological Science, Wood and colleagues had people repeat a grand positive self-statement, "I am a lovable person." For people with high self-esteem it helped a little. For people with low self-esteem it made them feel worse than those who did not repeat it at all. The authors concluded that broad, hard-to-believe statements can backfire for the very people who most want them to help.

It is worth being precise about what this does and does not mean. That backfire study used mantra-style repetition of a grand claim, which is not the same as the values-writing paradigm behind the education and health findings, and later work by Flynn and Bordieri in 2020 could not reproduce the effect, so it is a boundary condition rather than a law. The practical lesson holds either way: a line you cannot yet believe is more likely to sting than to soothe, which is why graded, specific wording matters.

What affirmations can and cannot do

Read together, the evidence points to a modest, conditional tool. Affirmations can lower defensiveness, buffer stress at a vulnerable moment, and make people more open to a message they would otherwise resist. They work best when they are believable, aimed at a real situation, and matched to the person and the moment, and they work alongside concrete action rather than instead of it.

They are not a treatment. No credible study shows that affirmations cure or replace care for anxiety, depression, or any clinical condition, and the honest reading of the health research is that they nudge intentions and reduce resistance, complementing action rather than substituting for it. The takeaway is neither hype nor dismissal: used well, affirmations are a small, low-cost support with real but bounded effects, not a standalone fix.