Self-Confidence Is Two Things, and the Evidence Favors Structured Methods

Researchers split self-confidence into worth (self-esteem) and capability (self-efficacy). Meta-analyses give the strongest support to structured approaches such as cognitive behavioral therapy and self-compassion training, though the evidence has real limits. Repeating positive statements can make people with low self-esteem feel worse.

Self-confidence is not a single trait in the research literature, and advice that strengthens one part of it may do little for another. Psychologists separate a general sense of personal worth, called self-esteem, from a task-specific belief in one's own ability, called self-efficacy. A third idea, self-compassion, is treated as related but distinct. The question of how to build confidence therefore depends on which of these is being built.

The best-supported methods in the research brief are structured ones: cognitive behavioral therapy (CBT) aimed at low self-esteem, and training in self-compassion. A popular shortcut, repeating upbeat statements about oneself, fares worse. In one set of experiments, it left people with low self-esteem feeling worse rather than better.

This matters because confidence advice is everywhere, and much of it assumes that feeling better about oneself automatically leads to better results. The evidence is more cautious than that. It also contains open disagreements and gaps, which this article lays out alongside what the studies found.

Confidence is at least two things

In everyday speech, confidence blends two separate ideas. Self-esteem describes how much a person feels worthy, in general. Self-efficacy describes whether a person believes they can handle a specific demand, such as giving a presentation or finishing a workout routine. One source in the research brief draws exactly this line: self-esteem is perceived self-worth, while self-efficacy is perceived capability.

Albert Bandura's theory of self-efficacy names four sources of that capability belief: mastery experiences (succeeding at something), vicarious experiences (watching others succeed), verbal persuasion (encouragement from others), and physiological and affective states (how the body and mood feel during a task). According to a paper on preservice teachers indexed in PubMed Central, the theory holds that mastery experiences generally exert the strongest pull, because they are the most authentic evidence of what a person can do.

Self-compassion, meaning a kind rather than punishing stance toward one's own failures, is the third construct. Researchers treat it as distinct from self-esteem. That distinction shapes how to read the trial evidence below, because the self-compassion studies measure self-compassion and self-criticism, not confidence itself. Any link to confidence is an inference.

How much self-esteem changes across a life

Before asking whether confidence can be raised, it helps to know how fixed it is. Ulrich Orth and Richard Robins, in a 2014 review in Current Directions in Psychological Science, describe self-esteem as a relatively stable trait that is by no means immutable. People who rank high or low at one stage of life tend to rank similarly decades later, yet the trait still moves.

The average direction of that movement has been mapped twice. A 2010 analysis by Orth, Trzesniewski and Robins used the US Americans' Changing Lives study and found a curved trajectory: self-esteem rose through young and middle adulthood, peaked at about age 60, and declined in old age. A later meta-analysis by Orth, Erol and Luciano, published in 2018 in Psychological Bulletin, covers ages 4 to 94. According to Orth's own summary of that work, self-esteem rises in childhood, holds roughly constant in early and middle adolescence, increases strongly between ages 15 and 30, and keeps climbing until about 60.

Two cautions apply. The excerpt of the summary available for this article cuts off after describing stability in the early 70s, so the late-life portion is better anchored in the 2010 cohort study. The same research group also reports that gender differences in self-esteem development are small, though the size and direction of the gap could not be confirmed here. These are averages, and averages can mask large differences between individuals.

The practical reading is modest. Confidence tends to grow on its own through much of adulthood, which means it is not fixed at birth, but this age pattern says nothing about whether any deliberate technique speeds the process.

Does high self-esteem actually cause success?

A common assumption holds that people who feel good about themselves will perform better, build stronger relationships and live healthier lives. A 2003 review by Roy Baumeister, Jennifer Campbell, Joachim Krueger and Kathleen Vohs, published in Psychological Science in the Public Interest, concluded that the findings do not support broad programs to raise self-esteem in the hope that this alone will improve outcomes. The authors also warned that praise handed out indiscriminately could just as easily foster narcissism.

Their alternative was to tie praise to behavior: reward socially desirable conduct and self-improvement rather than offering approval unconditionally. The review also found that self-esteem is strongly related to happiness. A secondary text in the brief says the causal direction of that relationship remains uncertain, so happy people may simply report higher self-esteem, or the reverse.

A competing reading from longitudinal work

Orth and Robins state in their 2014 review that high self-esteem prospectively predicts success and well-being in areas such as relationships and work. That sits uneasily beside the 2003 conclusion. The two are not necessarily contradictory, because predicting an outcome is not the same as causing it, and the 2003 review predates much of the longitudinal research. The brief's author could not read the later papers' methods for causal inference, so the question is best treated as unresolved.

For readers, the implication is that confidence may be worth building for its own sake, but there is little basis for expecting that a boost in self-regard will, by itself, lift performance.

What the trials support

Cognitive behavioral therapy for low self-esteem

CBT has the strongest meta-analytic backing in the brief for raising self-esteem. Kolubinski and colleagues published a systematic review and meta-analysis in Psychiatry Research in 2018 of CBT built on the Fennell model of low self-esteem. Participants were adults with low self-esteem and no severe, long-standing mental health condition or cognitive deficit. The effect size was not visible in the available excerpts, though a later paper describes the pooled effect as large; that is a secondhand description that needs checking against the original.

A 2021 meta-analysis of self-esteem interventions in adults, by Niveau and colleagues, is summarized by a later paper as finding that CBT, reminiscence-based interventions and evaluative conditioning all raised self-esteem, with the largest effects for CBT. A 2025 meta-analysis of brief group CBT pooled 31 studies with 6,109 participants and found improvements in depression, anxiety, self-esteem and overall mental health. Self-esteem was only one of several outcomes there, and the populations varied widely.

Research in this area continues. A registered trial, NCT06866639, plans to compare metacognitive therapy with CBT for low self-esteem, with 10 participants in each arm. Its documents describe a protocol, not results, and a study that small could at most offer preliminary signals.

Self-compassion training

Ferrari and colleagues, writing in Mindfulness in 2019, meta-analyzed 27 randomized controlled trials of self-compassion interventions and reported improvements across many psychosocial outcomes. Kristin Neff's 2023 chapter in the Annual Review of Psychology summarizes that work as showing medium to large gains in self-compassion and reductions in psychopathology, often from short interventions. Neff is a leading developer of self-compassion research and of an intervention program, and both documents are hosted on her own website, so independent confirmation carries extra weight.

The cleanest numbers come from Wakelin and colleagues, who pooled randomized trials in Clinical Psychology & Psychotherapy in 2022. Among 20 trials, 19 papers with 1,350 participants had enough data. Compared with control groups, self-compassion interventions produced a significant, medium reduction in self-criticism (Hedges' g = 0.51, 95% confidence interval 0.33 to 0.69). Longer interventions were associated with larger effects, and the authors recommended more trials with active control groups.

The outcomes in these trials are self-compassion and self-criticism. Whether less self-criticism translates into more confidence is a reasonable inference, not a demonstrated result.

Mastery experiences and self-efficacy

For capability beliefs, Bandura's theory ranks success experiences first. The test of that ranking is mixed. A 2025 meta-analysis of interventions to raise teachers' self-efficacy, by Täschner and colleagues in Review of Educational Research, found that programs including mastery experiences did not differ significantly overall from those without. For pre-service teachers specifically, programs offering only mastery experiences produced the largest self-efficacy gains.

That result covers one occupation and cannot be generalized to confidence broadly. A separate 2025 paper on physical activity, indexed in PubMed Central, notes that researchers still debate whether some sources of self-efficacy matter more than others. Mastery remains the best-theorized route, with only partial empirical confirmation.

When positive self-talk backfires

The most familiar confidence technique is repeating a flattering statement until it sinks in. Joanne Wood, W. Q. Elaine Perunovic and John Lee tested this in a survey and two experiments, reported in Psychological Science in 2009. The survey found that people commonly use positive self-statements and believe they work.

The experiments told a different story. Among participants with low self-esteem, those who repeated "I'm a lovable person" felt worse than those who did not repeat it or who thought about how the statement was both true and not true. Among participants with high self-esteem, repeating it improved their mood, but only to a limited degree. In other words, the technique appeared to help least the people most likely to reach for it.

The evidence here is thin. The brief's author saw the study only through its abstract and secondary coverage, found no replication, and notes that participants appear to have been students and that the effects were measured on immediate mood and state self-esteem rather than lasting confidence. The finding fits with the 2003 review's caution about unearned praise, but a single research group's experiments are suggestive, not settled.

What the evidence supports, and what remains open

Several limits run through the material. The CBT meta-analysis was restricted to one therapy model and to adults without severe disorders, so it may not carry over to the general public or to self-help use. Most self-compassion trials are short, many lack active controls, and outcomes are self-reported. The lifespan data describe averages, and factors such as generational differences and participants dropping out could affect them, though this was not verified.

Taken together, the studies suggest that confidence is more likely to be built by structured, repeated work than by slogans. That work includes addressing the self-critical thinking that sustains low self-esteem, treating failures with less harshness, and accumulating real evidence of capability. Praise appears to work best when it is attached to effort and improvement.

What to watch is the kind of study that could settle the open questions: large, pre-registered randomized trials with active comparison groups, long follow-up, and behavioral outcomes as well as questionnaires. Until then, the findings are best read as moderate-confidence guidance, researched context rather than personal or clinical advice.

Frequently asked questions

What is the difference between self-esteem and self-confidence?

Self-confidence is a lay term that research splits into self-esteem, a general sense of worth, and self-efficacy, belief in one's capability for a specific task. The two respond to different influences, so advice for one may not help the other.

Does self-esteem change as people get older?

Yes, on average. A 2010 US longitudinal analysis found self-esteem rose through young and middle adulthood, peaked at about age 60, and declined in old age. A meta-analysis covering ages 4 to 94 describes a strong increase from age 15 to 30. Individual paths vary.

Do positive affirmations work for low self-esteem?

A 2009 set of experiments found that people with low self-esteem who repeated "I'm a lovable person" felt worse afterward than those who did not. People with high self-esteem felt somewhat better. The evidence comes from one research group, and no replication was found for this article.

What has the best evidence for raising self-esteem?

Among the sources reviewed, CBT designed for low self-esteem has the strongest meta-analytic support, though effect sizes could not be confirmed from the original papers. Self-compassion training shows a medium reduction in self-criticism in randomized trials, but that is not the same as higher confidence.

Does building self-esteem lead to better performance and success?

The evidence is contested. A 2003 review found little support for boosting self-esteem alone to improve outcomes. Later longitudinal work reports that high self-esteem predicts success in relationships and work, but prediction does not prove cause.

Sources

1. Self-Esteem Development From Young Adulthood to Old Age: A Cohort-Sequential Longitudinal Study , Journal of Personality and Social Psychology (American Psychological Association)
2. The Development of Self-Esteem , Current Directions in Psychological Science
3. Development of Self-Esteem From Age 4 to 94 Years: A Meta-Analysis of Longitudinal Studies , Psychological Bulletin
4. Self-Esteem (book chapter) , Ulrich Orth, University of Bern
5. Does High Self-Esteem Cause Better Performance, Interpersonal Success, Happiness, or Healthier Lifestyles? , Psychological Science in the Public Interest
6. A systematic review and meta-analysis of CBT interventions based on the Fennell model of low self-esteem , Psychiatry Research
7. Self-esteem Interventions in Adults: A Systematic Review and Meta-analysis , ResearchGate listing
8. Effectiveness of brief group cognitive behavioural therapy on improving mental health outcomes , ScienceDirect
9. Cognitive Behavioural Therapy/Metacognitive Therapy for Low Self Esteem (statistical analysis plan, NCT06866639) , ClinicalTrials.gov
10. Self-Compassion Interventions and Psychosocial Outcomes: A Meta-Analysis of RCTs , Mindfulness
11. Self-Compassion: Theory, Method, Research, and Intervention , Annual Review of Psychology
12. Effectiveness of self-compassion-related interventions for reducing self-criticism: A systematic review and meta-analysis , Clinical Psychology & Psychotherapy (Wiley)
13. Why Do I Feel More Confident? Bandura's Sources Predict Preservice Teachers' Latent Changes in Teacher Self-Efficacy , PubMed Central (Frontiers)
14. "Yes, I Can!" A Systematic Review and Meta-Analysis of Intervention Studies Promoting Teacher Self-Efficacy , Review of Educational Research
15. An empirical ranking of the importance of the sources of self-efficacy for physical activity , PubMed Central
16. Positive Self-Statements: Power for Some, Peril for Others , Psychological Science (via Semantic Scholar)

This article was drafted with the help of AI and reviewed by Jaroha editors before publication.