Self-esteem statistics vary widely because studies use different scales, cutoffs, countries, age groups, and samples. The figures below show measured patterns in Brazil, India, the United States, Japan, Vietnam, the United Arab Emirates, and Iran, with each study’s dates and definitions kept explicit.
Contents
- How common are low and high self-esteem scores?
- How self-esteem changes from childhood into adolescence
- Adolescent benchmarks by age and sex
- U.S. gender and race or ethnicity patterns
- What Japanese surveys show about measurement
- Youth self-evaluation and well-being indicators
- What intervention studies measured
How common are low and high self-esteem scores?
A 2017 cross-sectional study of 1,011 Brazilian adolescents found that 33.8% had low self-esteem on the Rosenberg scale. When moderate scores were included, 64.3% had either moderate or low self-esteem. These percentages describe that study sample and its classification method; they are not a prevalence estimate for all Brazilian adolescents. Influence of Body Dissatisfaction on the Self-Esteem of Brazilian Adolescents: A Cross-Sectional Study
The same Brazilian sample showed a different distribution by sex:
| Group | Low | Moderate | High |
|---|---|---|---|
| Girls | 34.3% | 29.3% | 36.4% |
| Boys | 37.4% | 31.6% | 31.0% |
The largest difference in these reported categories was the share classified as high: 36.4% among girls compared with 31.0% among boys. The study was measured in 2017 and used a cross-sectional design, so the figures describe associations and distributions at one period rather than changes over time. Brazilian adolescent study
A separate 2023 study of 60 adolescents in Visnagar, India, reported a much more concentrated low-score distribution before assertiveness training. Seventy percent were classified with low self-esteem, 30.0% with average self-esteem, and 0% with high self-esteem. After the program, 63.3% were classified with high self-esteem, 36.7% with average self-esteem, and 0% with low self-esteem. Because this was a small intervention sample, the before-and-after percentages should be read as results for that group, not as a population benchmark. Data on self-esteem among adolescents in India
How self-esteem changes from childhood into adolescence
A 2021 child-scale study reported mean scores and the percentage below its low-self-esteem cutoff for children aged 7–8, 9–10, and 11–12. The scale and cutoff matter: these numbers should not be directly compared with results from a different Rosenberg version or scoring range.
| Age and sex | Mean score | 95% confidence interval | Below low cutoff |
|---|---|---|---|
| Boys, 7–8 | 31.78 | 31.00–32.56 | 1.9% |
| Boys, 9–10 | 28.65 | 27.82–29.48 | 7.9% |
| Boys, 11–12 | 29.63 | 28.55–30.71 | 4.8% |
| Girls, 7–8 | 31.35 | 30.56–32.13 | 1.9% |
| Girls, 9–10 | 29.59 | 28.69–30.49 | 5.0% |
| Girls, 11–12 | 28.75 | 27.34–30.11 | 3.8% |
In this study, the reported mean was highest for boys at ages 7–8 and lowest for boys at ages 9–10 among the three male age groups. For girls, the mean declined at each successive age band, from 31.35 at ages 7–8 to 28.75 at ages 11–12. The percentage below the low cutoff was lowest at ages 7–8 for both sexes, at 1.9%, and higher in the older groups.
The study found a statistically significant age effect, F(2,711)=20.32, p<0.001, with partial eta squared of 0.055. It also reported that self-esteem was significantly lower at ages 9–10 and 11–12 than at ages 7–8, with p<0.001 for both comparisons. Modification of the Rosenberg Scale to Assess Self-Esteem in Children
These results illustrate why an age-specific statistic is more useful than a single number for all young people. The study measured a developmental pattern in its child sample, while the exact score differences depend on the instrument, sample, and cutoff used.
Adolescent benchmarks by age and sex
The same source reports published adolescent norms from 2001. These are historical norms, not a newly collected 2026 benchmark. They provide reference means for adolescents aged 12–13 through 18–19:
| Age | Male mean | Female mean |
|---|---|---|
| 12–13 | 30.18 | 28.51 |
| 14–15 | 31.16 | 27.86 |
| 16–17 | 30.54 | 28.37 |
| 18–19 | 31.68 | 28.84 |
Across these published norms, male means ranged from 30.18 to 31.68, while female means ranged from 27.86 to 28.84. The male-minus-female gap was 1.67 points at ages 12–13, 3.30 points at ages 14–15, 2.17 points at ages 16–17, and 2.84 points at ages 18–19. Those differences are arithmetic descriptions of the displayed norms, not evidence that every individual follows the group pattern. Modification of the Rosenberg Scale to Assess Self-Esteem in Children
The contrast between the child-scale results and these adolescent norms also reinforces a measurement caution. A mean score only has meaning alongside the scale version, score range, reference population, and measurement year. A higher or lower number cannot be interpreted in isolation across studies.
U.S. gender and race or ethnicity patterns
A U.S. study covering survey years 1991–1993 and 2006–2008 reported small changes in mean self-esteem among female students. Among 8th-grade females, the mean rose from 3.88 to 3.96, with Cohen’s d=0.09. Among 10th-grade females, it rose from 3.90 to 3.97, with d=0.07. Among 12th-grade females, it rose from 4.00 to 4.04, with d=0.05.
The pattern for 12th-grade males was slightly different: the mean declined from 4.12 in 1991–1993 to 4.08 in 2006–2008, with Cohen’s d=-0.05. The study estimated that no more than 0.25% of total self-esteem-score variance was between survey years, and that linear time trends accounted for 0.08% or less of total variance. These estimates indicate that the reported time-period differences were small relative to overall score variation. Adolescent Self-Esteem: Differences by Race/Ethnicity, Gender, and Age
The same U.S. analysis reported a race or ethnicity pattern within 12th-grade females: 45% of African-American females were in the four highest self-esteem index categories, compared with 24% of Asian-American females. These category shares describe the study’s index and groups during the covered survey period; they should not be generalized to all students or treated as explanations for the difference. U.S. adolescent self-esteem study
What Japanese surveys show about measurement
Six Japanese surveys conducted from 2009 through 2018 examined the reliability and relationships among parts of the Rosenberg self-esteem scale across adulthood. Reliability for the full scale was above alpha=0.86 in every survey. The self-competence subscale had reliability above alpha=0.77, while the self-liking subscale exceeded alpha=0.69.
The two subscales were strongly related in the reported samples. Self-competence and self-liking correlated above r=0.72 for the total Japanese samples, above r=0.68 among Japanese males, and above r=0.72 among Japanese females. At the age-group level, the correlation exceeded r=0.71 among Japanese males and r=0.79 among Japanese females. The Developmental Trajectory of Self-Esteem Across the Life Span in Japan
These are measurement statistics rather than percentages of people with low self-esteem. Reliability describes consistency of a scale, while correlation describes how two measured components move together. Keeping those meanings separate helps prevent a reliability coefficient from being mistaken for a population score.
Youth self-evaluation and well-being indicators
The SAVY 2 survey, reported for Vietnam in 2010, found an average self-esteem score of 2.76. Among youth aged 14–17, the average was 2.68. The survey also reported responses to several self-evaluation statements, with male and female percentages shown separately.
| Statement completely agreed with | Male youth | Female youth |
|---|---|---|
| Had some good qualities | 65.5% | 67.5% |
| Were not proud of themselves | 20.9% | 23.5% |
| Could do what others did | 65.8% | 66.6% |
| Sometimes felt they were no good at all | 28.7% | 31.1% |
| Were valuable to their families | 78.8% | 75.9% |
The table shows that positive and negative self-evaluations can coexist in the same survey. For example, 65.5% of male youth and 67.5% of female youth completely agreed that they had some good qualities, while 28.7% and 31.1%, respectively, completely agreed that they sometimes felt they were no good at all. Health of Adolescents in Viet Nam
In a UAE adolescent school sample measured in 2017–2018, depressive symptoms were detected in 17.2% of participants, with a 95% confidence interval of 14.2%–20.7%. The study reported that depressive symptoms were inversely related to self-esteem scores. This is an association in that sample, not evidence that the self-esteem measure alone caused or prevented depressive symptoms. Self-esteem and other risk factors for depressive symptoms among adolescents in United Arab Emirates
What intervention studies measured
Intervention results can show measured change within a group, but they do not automatically establish how much change would occur in every setting. In the 2023 Indian assertiveness-training study, the mean self-esteem score increased from 11.33 before training to 21.16 after training. The reported mean pre/post difference was 9.83 points, with a paired t value of 33.4. The pre-test mean standard deviation was 1.2 points, and the post-test mean standard deviation was 1.9 points. Data on self-esteem among adolescents in India
Among female secondary-school students in Shiraz, Iran, a 2014 peer-led intervention group had a mean self-esteem score of 51.80 (SD 13.91) before training and 73.72 (SD 12.94) at the first post-test. At the second post-test, six weeks after training, the mean was 69.48 (SD 12.63). The second post-test mean was lower than the first post-test mean, while remaining above the pre-training mean. The effects of a peer-led training program on female students’ self-esteem in public secondary schools in Shiraz
A 2019 middle-school health-education study of male students reported baseline means of 28.37 (SD 3.58) in the intervention group and 27.89 (SD 3.82) in the control group. Immediately after education, the intervention mean was 32.17 (SD 3.12), compared with 27.50 (SD 3.56) in controls. Three months later, the intervention mean was 33.83 (SD 3.32), compared with 27.32 (SD 4.37) in controls. The Effects of Multimedia-Based Puberty Health Education on Male Students’ Self-Esteem in the Middle School
Taken together, these intervention studies report increases in measured self-esteem in particular adolescent groups, using different instruments, score ranges, designs, and follow-up periods. The results are best used as study-specific evidence about the measured programs and participants, not as a universal expected increase.