Self-esteem test accuracy: what research says and why your score changes

Your self-esteem test gives 14, “A season of being hard on yourself”; a few weeks later you take it again and get 16, “Mostly steady self-esteem”; or another site gave you 28 out of 40 and called it “normal.” Stories like these are common, and they raise the obvious questions: which one is right, and can a self-esteem test be trusted at all?

The short answer: the Rosenberg Self-Esteem Scale (Rosenberg, 1965) that mind.io uses has measured self-esteem fairly consistently in research. The University of Maryland’s Department of Sociology, which hosts the scale’s guidance pages, summarises its internal consistency across samples as mostly .77 to .88 and its retest correlations as .82 to .88 (University of Maryland, n.d.-a). The same department says there are no fixed cut-off points that separate high from low self-esteem (University of Maryland, n.d.-b). The same scale can also give different numbers depending on the translation, the wording, the answer options and the scoring, and mind.io asks you to answer with recent weeks in mind, so how things have gone lately shows up in the score too.

This guide covers how reliable the Rosenberg scale has been in research, why there is no official score below which self-esteem counts as low, the debate over its positive and negative statements, why your score moves when you take the test again, what a self-esteem score can miss, and how to read and use your result. Where your score on the mind.io self-esteem test sits next to Korean adults is covered in average self-esteem score, the patterns that come with low self-esteem in signs of low self-esteem, and habits that build it in how to build self-esteem.

The Rosenberg Self-Esteem Scale is a 10-item scale created by the sociologist Morris Rosenberg. It was developed in the 1960s with high school juniors and seniors from 10 randomly selected schools in New York State, 5,024 in all, and was first scored as a Guttman scale, in which items stack like steps; today the answers, usually one of four options for each statement, are simply added up (University of Maryland, n.d.-a). What it aims to measure has been described as feeling “good enough” rather than feeling “very good” compared with others (Rosenberg, 1965, as cited in Uchida & Ueno, 2010). The accuracy of a scale like this is usually judged three ways: whether the items point the same way (internal consistency), whether a retest after a short gap gives similar results (test-retest reliability), and whether it measures what it sets out to measure and relates sensibly to other measures (validity).

Research has generally supported the scale. The Maryland department summarises internal consistency across samples as mostly .77 to .88 and retest correlations as .82 to .88 (University of Maryland, n.d.-a). In a study that gave the scale, translated into 28 languages, to people in 53 nations, 16,998 in all, the way the items grouped was largely the same across nations, and in nearly all of them scores were related to neuroticism, extraversion and romantic attachment styles (Schmitt & Allik, 2005, from the abstract). Another study found that a one-item self-esteem measure showed nearly the same pattern of correlations as the Rosenberg scale in adults, though the two agreed only moderately in children (Robins et al., 2001, from the abstract).

In Korea, several translations have been used, including one from 1974. In a study of 3,498 students (elementary, middle and high school students in and around Seoul, and university students from several regions), internal consistency was .75 to .87, but item 8, translated as “I wish I could respect myself a little more” (reverse-scored), behaved differently from the rest. While the other items correlated .43 to .71 with the total, item 8 correlated −.23 in elementary, −.18 in middle and −.02 in high school students and .14 in university students, and without it internal consistency rose to .81, .85, .83 and .89. The authors suggested that the English “I wish …” may have read in Korean as a hope to grow rather than as a sense of falling short (Lee et al., 2009a). A follow-up by the same team compared three revised wordings in 338 middle and high school students and found that “I regret that I cannot respect myself more” fitted best (Lee et al., 2009b, from the abstract), and an analysis of 2,351 first-year middle school students in the Korean Children and Youth Panel Survey, which had kept the original item 8, left it out and used nine items (Choi & Cho, 2013). In the 53-nation study, too, 10 nations had a loading of .30 or less on item 8 (Schmitt & Allik, 2005, as cited in Lee et al., 2009a).

In Japan, an analysis of studies published in journals from 1980 to March 2013, 256 of them with 48,927 people in all, found several translations in use: Yamamoto et al.’s (1982) in 59.8% and Hoshino’s (1970) in 15.2%, while only 20.3% of the studies used the four answer options Rosenberg used (Oshio et al., 2014). Yamamoto et al. (1982), the most used, found one factor in data from 644 university students and then added up nine items, without saying which item was left out (Yamamoto et al., 1982; Uchida & Ueno, 2010). A Japanese version refined by translating back and forth with the English (the RSES-J) showed nearly the same structure as the original in 1,320 Japanese speakers (Mimura & Griffiths, 2007, from the abstract); in 329 university students its internal consistency was .81, and in the 50 of them who answered again two weeks later the retest agreement (ICC) was .78. In this version too, item 8 had the lowest loading of the ten items, at .40, but it did not run the opposite way (Uchida & Ueno, 2010).

In French, a French-Canadian version was checked in four studies: internal consistency (study 1) and a single-factor structure (study 2) in junior college students, then construct validity (study 3) and stability over time (study 4), though the authors called their findings preliminary (Vallières & Vallerand, 1990, from the abstract). A study of the French version in older adults found one global self-esteem factor, with method effects tied more to the negatively worded statements (Gana et al., 2005, from the abstract).

The mind.io self-esteem test asks the scale’s 10 items with four answers from “Strongly agree” to “Strongly disagree,” scores them 3 to 0, reverses the five statements that put you down (items 2, 5, 6, 8 and 9) and adds them up to a score from 0 to 30. The wording in Korean, English, Japanese and French is mind.io’s own translation or rewording, not the versions in the studies above, and mind.io has not studied the reliability or validity of its test. So the figures above describe research versions of the scale, not how accurate a mind.io result is.

Depression and anxiety self-checks set their cut-offs by comparing scores with diagnostic interviews: how many people with the condition a cut-off finds (sensitivity) and how many without it it clears (specificity). Self-esteem is not a diagnosis, so there is nothing of the kind to check a score against. The Maryland department itself says there are no discrete cut-off points for high and low self-esteem and advises looking for norms in studies of people like those you want to compare with (University of Maryland, n.d.-b). The averages for Korean adults are in average self-esteem score.

So the lines vary with whoever draws them. In Japan, a 1984 study noted that young people averaged around 25 on four-option translations (scored 10 to 40), and it also proposed, as a very simple rule of thumb, reading 20 or less as low and 30 or more as high (as cited in Uchida & Ueno, 2010). On mind.io’s 0 to 30 scoring that is calculated as 10 or less and 20 or more. The numbers themselves can also differ for the same scale. Scoring the four options 0 to 3, as the Maryland department’s instructions do, gives 0 to 30, scoring them 1 to 4 gives 10 to 40, and some researchers use five- or seven-point answers (University of Maryland, n.d.-a). In the analysis of 256 Japanese studies, studies with five-point answers had significantly, if slightly, higher means than the others, while the translation used made no significant difference (Oshio et al., 2014).

Item numbers differ between versions too. In the version the Maryland department publishes, mind.io’s item 7 (a person of worth, at least as much as others) is item 1 and mind.io’s item 1 (satisfied on the whole) is item 7, though the statement about wishing for more self-respect is item 8 in both (University of Maryland, n.d.-a). So before comparing with a score from another site, check that the answer options and scoring match. A score out of 40 from four-option answers can be compared with mind.io’s after subtracting 10.

mind.io’s result also splits scores into 0–14, “A season of being hard on yourself,” 15–25, “Mostly steady self-esteem,” and 26–30, “Solid self-esteem,” but these ranges are mind.io’s reading aids, not cut-offs set by research. The middle range is wide, so 15–20 and 21–25 get different paragraphs. A 14 and a 15 are not two kinds of people; the names are short labels for positions on a continuous score.

Of the Rosenberg scale’s 10 items, five affirm you and five put you down. With the two mixed, a habit of agreeing regardless of content cannot push the total in one direction only. Calculated with mind.io’s scoring, giving the same answer to every statement, whether “Strongly agree” or “Strongly disagree,” always gives a total of 15.

How to read the two strands has long been debated. In the Korean Welfare Panel analysis, the data fit better with the ten items split into two strands than read as one (average self-esteem score). One reading is that they are two different kinds of self-esteem; the other is that they are one self-esteem overlaid with a difference in how people answer negative statements (a method effect). Responses from US adolescents reflected a relatively unidimensional self-esteem factor plus method effects tied to negative statements (Marsh, 1996, from the abstract), and another analysis also found a global self-esteem factor but needed method effects for a good fit (Tomás & Oliver, 1999, from the abstract). In adolescents measured four times over 8 years, support for the two-factor model weakened, and the method effects were not passing artefacts but fairly stable answering habits (Marsh et al., 2010, from the abstract).

A meta-analysis of 113 samples with 140,671 people found that a model with a global self-esteem factor plus factors for positive and negative statements fitted best, but the global factor took most of the explained common variance and the two strands less than 15%. The items’ ties to the global factor were the same in the US and other highly individualistic nations but weaker in less individualistic ones, so the authors doubted that scores can be compared across nations as they are (Gnambs et al., 2018, from the abstract). In data from 12,437 German students, the scale was closer to a single factor among those with stronger reading and reasoning skills, and the authors read the wording effect as an answering style linked to cognitive ability (Gnambs & Schroeders, 2020, from the abstract). In Korean first-year middle school students, too, a general factor plus a method factor for negative items fitted best (Choi & Cho, 2013). That does not mean the total is useless: in 1,127 college students, the global factors from the structural models that fitted related to the criteria much as the total score did, which the authors saw as reducing, but not eliminating, concerns about total scores (Donnellan et al., 2016, from the abstract).

On the mind.io result page, “Self-regard” (how much you agreed with the five positive statements) and “Self-critical voice” (how much you agreed with the five negative ones) are a reference split, not official subscales. Calculated with mind.io’s scoring, answering “Agree” to every statement gives 10 and 10, and “Strongly agree” to every one gives 15 and 15, which the page reads as both voices being present at once. “Disagree” everywhere gives 5 and 5, and “Strongly disagree” everywhere 0 and 0, which it reads as a cautious answer that also holds back from affirming your good qualities. So before reading much into the gap between the two, ask yourself whether a habit of answering one way, rather than the statements themselves, crept in. If the self-critical voice is loud, self-compassion versus self-esteem may help too.

In research, self-esteem shows up as a fairly stable trait. The Japanese version gave a retest agreement of .78 in 50 university students who answered again two weeks later (Uchida & Ueno, 2010), and the retest correlations the Maryland department summarises are .82 to .88 (University of Maryland, n.d.-a). A study that pooled 50 published articles with 29,839 people and four large national datasets with 74,381 found that the order of people’s scores held up to a degree, corrected for measurement error, in the .50s to .70s, comparable to personality traits; this stability was low in childhood, rose through adolescence and young adulthood and fell again from midlife (Trzesniewski et al., 2003, from the abstract). In 3,180 people aged 14 to 102 measured six times over 29 years, the correlation fell as the gap grew but levelled off near .43, and the authors saw self-esteem as trait-like (Kuster & Orth, 2013, from the abstract). How the average changes with age is in signs of low self-esteem.

A single score still carries your recent state. The original instructions ask about your general feelings about yourself (University of Maryland, n.d.-a), but mind.io asks you to answer with recent weeks in mind, so a hard few weeks naturally shows in the score. When an earlier result is saved in the same browser, the result page shows how many points it changed and notes that self-esteem scores naturally rise and fall with your energy, relationships and how things are going, so it is the trend that counts, not one change.

Calculated with mind.io’s scoring, changing one answer by one step moves the total by exactly 1 point. So 14, “A season of being hard on yourself,” and 15, “Mostly steady self-esteem,” and 25, “Mostly steady self-esteem,” and 26, “Solid self-esteem,” are one answer apart, and the paragraphs for the lower (15–20) and upper (21–25) part of the middle range split between 20 and 21. The result page does not flag a score near a range edge, so look at the number rather than the range name.

One statement needs particular care. mind.io’s Korean item 8, its translation of “I wish I could respect myself more,” is worded almost exactly like the Korean item that behaved differently from the rest in the study above, and mind.io has not changed it. It is reverse-scored, from 3 for “Strongly disagree” to 2, 1 and 0 for “Strongly agree,” so calculated, it is worth 0 to 3 of the 30 points. If you answered “Agree” (1 point) out of a wish to respect yourself more rather than a sense of falling short, your total may have come out 1–2 points lower than if you had read the statement as a shortfall (up to 3 if you chose “Strongly agree”). That is mind.io’s reading of the research; it is not from a study of mind.io’s users. In the Japanese version, item 8 was weakly tied to the rest but did not run the opposite way, and we did not find a study showing a problem with this item alone in English or French.

A self-esteem test asks you to rate how you see yourself, so a high score can cover many things. Psychologists who reviewed the research noted that people who report high self-esteem include those who frankly accept their good qualities along with narcissistic or defensive people, and that many people with high self-esteem exaggerate their successes and good traits; their claims of being more likable and having better relationships were mostly not borne out by objective measures (Baumeister et al., 2003, from the abstract). That does not make self-esteem unimportant: a review of studies that followed people over time concluded that high self-esteem predicts good outcomes in relationships, work and health (Orth & Robins, 2014, from the abstract). The difference between self-esteem and narcissism is also covered in narcissism.

Be careful comparing countries too. In the 53-nation study, every nation’s average was above the scale’s theoretical midpoint, but people in more collectivist cultures tended to answer toward the middle, and negative statements were read differently from nation to nation, which limits direct comparisons (Schmitt & Allik, 2005, from the abstract). Japan’s average was reported as the lowest in that study (Schmitt & Allik, 2005, as cited in Oshio et al., 2014). Some argue that a self-critical focus is more characteristic of Japanese culture (Heine et al., 1999, from the abstract), while another study found that self-esteem measured by questionnaire was lower in Japan but not on a test of automatic, implicit responses (Yamaguchi et al., 2007, as cited in Oshio et al., 2014). Your own score is better compared with similar people measured on the same version than with another country’s average.

A self-esteem score cannot pick out depression either. Studies link low self-esteem with later depression (see signs of low self-esteem), but the mind.io self-esteem test does not diagnose depression, anxiety or other mental health difficulties. If a low mood has lasted more than two weeks, take the depression check-in (PHQ-9) as well. That is not a diagnosis either; how to read its score is in PHQ-9 accuracy.

The Rosenberg scale asks only how you see yourself overall, so it does not show self-esteem that differs by area, such as work, relationships or looks. And do not use the score to judge other people: asking a child, partner or colleague to take the test and labelling them “someone with low self-esteem,” or using it in hiring or appraisals, is not what this test is for.

First, look at the number and its distance from the range edges rather than at the range name. 14 and 15, or 25 and 26, are one answer apart: different names, same neighbourhood.

Second, when you look at the two direction scores, think about whether you answered one way regardless of the statements, and how you read item 8.

Third, compare like with like. On mind.io’s scale, Korean Welfare Panel adults averaged about 18 to 22 (average self-esteem score). Subtract 10 from a score out of 40 from four-option answers before comparing, and do not compare directly with scores from five- or seven-point versions, other translations or another country’s average.

Fourth, if you retake the test, leave a few weeks and take it in similar circumstances. Retaking it straight away because you did not like the result makes it easy to remember the items and pick the answers you want. Note briefly what happened in between, and look at the direction over several results rather than a few points once. To turn it into habits, how to build self-esteem is a good place to start.

Fifth, if the way you see yourself has felt heavy for weeks, or your mood is so low that daily life is hard to manage, do not carry it alone: talk to someone you trust, a counselling service at school or work, or a professional counsellor. The mind.io result page also shows this kind of help note for scores of 0 to 14, and adds crisis line details for scores of 9 or below.

What not to do: do not use one score to label yourself “someone with low self-esteem,” treat a high score as proof that you are better than others, or judge a culture by comparing country averages. A self-esteem score is a gauge of how you see yourself lately, not a verdict.

Q. Can a self-esteem test be trusted? A. The research versions of the Rosenberg scale hang together fairly well and give similar results on a retest. The Maryland department summarises internal consistency as mostly .77 to .88 (University of Maryland, n.d.-a); it was .75 to .87 in Korean students (Lee et al., 2009a), and the Japanese version showed .81 with a two-week retest agreement of .78 (Uchida & Ueno, 2010). But there is no official line between high and low, and mind.io’s wording has not been validated separately, so use it as a rough gauge of how you see yourself lately.

Q. What score counts as low self-esteem? A. There is no fixed cut-off (University of Maryland, n.d.-b). mind.io’s 15 and 26 are only edges that make the result easier to read. On mind.io’s scale, Korean Welfare Panel adults averaged about 18 to 22, so see where your score sits in average self-esteem score.

Q. My score differs from another site’s. A. Versions differ in scoring, as with 0 to 30 versus 10 to 40, and some use five- or seven-point answers (University of Maryland, n.d.-a). In Japanese studies, five-point versions gave slightly higher means, while the translation made no significant difference (Oshio et al., 2014). Item order differs between versions too, so compare only scores with the same answer options and scoring.

Q. My score changed since last time. Which one is right? A. Both reflect you at the time. On mind.io a single different answer moves the total by 1 point, and even the research version in Japanese had a two-week retest agreement of .78 (Uchida & Ueno, 2010). mind.io asks you to answer with recent weeks in mind, so look at the direction over several results together with what happened in between.

Q. I was unsure about item 8. A. In the Korean student study, the Korean “I wish I could respect myself a little more” behaved differently from the other items (Lee et al., 2009a), and a study comparing revised wordings proposed “I regret that I cannot respect myself more” (Lee et al., 2009b, from the abstract). mind.io’s Korean item 8 means the same thing, and calculated, it is worth 0 to 3 points of the total. In mind.io’s reading, if you read it as a wish and answered “Agree” (1 point), your total may have come out 1–2 points lower than if you had read it as a shortfall (up to 3 if you chose “Strongly agree”).

Q. Do self-esteem scores differ by country? A. The 53-nation study found that negative statements were read differently from nation to nation, which limits direct comparisons (Schmitt & Allik, 2005, from the abstract). Whether Japan’s lowest average (Schmitt & Allik, 2005, as cited in Oshio et al., 2014) reflects answering habits or a real difference is debated. Compare yourself with similar people measured on the same version rather than with a country average.

Q. Is a high self-esteem score good? A. High scores mix people who frankly accept their good qualities with defensive or narcissistic people (Baumeister et al., 2003, from the abstract). Still, a review found that high self-esteem predicted good outcomes in relationships, work and health (Orth & Robins, 2014, from the abstract). What the Rosenberg scale aims at is closer to feeling “good enough” than to feeling better than others.

Q. Do I need counseling if my self-esteem is low? A. Low self-esteem is not a diagnosis, and much of it can be practiced on your own. Counseling can still help even when there is no crisis. But if a low mood lasts or you have thoughts of harming yourself, reach out to a professional or a crisis line right away: 988 in the US and Canada (call or text), Samaritans on 116 123 in the UK and Ireland, or findahelpline.com elsewhere. If you are in immediate danger, contact your local emergency services.

References: Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003), Does high self-esteem cause better performance, interpersonal success, happiness, or healthier lifestyles?, Psychological Science in the Public Interest 4(1), 1-44, https://doi.org/10.1111/1529-1006.01431; Choi, S., & Cho, Y. I. (2013), Measurement invariance of self-esteem between male and female adolescents and method effects associated with negatively worded items [in Korean], Korean Journal of Psychology: General 32(3), 571-589, https://www.accesson.kr/kpageneral/v.32/3/571/15898; Donnellan, M. B., Ackerman, R. A., & Brecheen, C. (2016), Extending structural analyses of the Rosenberg Self-Esteem Scale to consider criterion-related validity: Can composite self-esteem scores be good enough?, Journal of Personality Assessment 98(2), 169-177, https://doi.org/10.1080/00223891.2015.1058268; Gana, K., Alaphilippe, D., & Bailly, N. (2005), Factorial structure of the French version of the Rosenberg Self-Esteem Scale among the elderly, International Journal of Testing 5(2), 169-176, https://doi.org/10.1207/s15327574ijt0502_5; Gnambs, T., Scharl, A., & Schroeders, U. (2018), The structure of the Rosenberg Self-Esteem Scale: A cross-cultural meta-analysis, Zeitschrift für Psychologie 226(1), 14-29, https://doi.org/10.1027/2151-2604/a000317; Gnambs, T., & Schroeders, U. (2020), Cognitive abilities explain wording effects in the Rosenberg Self-Esteem Scale, Assessment 27(2), 404-418, https://doi.org/10.1177/1073191117746503; Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999), Is there a universal need for positive self-regard?, Psychological Review 106(4), 766-794, https://doi.org/10.1037/0033-295X.106.4.766; Kuster, F., & Orth, U. (2013), The long-term stability of self-esteem: Its time-dependent decay and nonzero asymptote, Personality and Social Psychology Bulletin 39(5), 677-690, https://doi.org/10.1177/0146167213480189; Lee, J., Nam, S., Lee, M., Lee, J., & Lee, S. M. (2009a), Rosenberg’s Self-Esteem Scale: Analysis of item-level validity [in Korean], The Korean Journal of Counseling and Psychotherapy 21(1), 173-189, https://accesson.kr/kcpa/v.21/1/173/20529; Lee, J., Nam, S., Lee, J., Choi, B., Park, Y., & Lee, S. M. (2009b), Errors in item translation of psychological assessment by cultural discrepancy: Revising 8th item of Rosenberg’s Self-Esteem Scale [in Korean], Korea Journal of Counseling 10(3), 1345-1358, https://doi.org/10.15703/kjc.10.3.200909.1345; Marsh, H. W. (1996), Positive and negative global self-esteem: A substantively meaningful distinction or artifactors?, Journal of Personality and Social Psychology 70(4), 810-819, https://doi.org/10.1037/0022-3514.70.4.810; Marsh, H. W., Scalas, L. F., & Nagengast, B. (2010), Longitudinal tests of competing factor structures for the Rosenberg Self-Esteem Scale: Traits, ephemeral artifacts, and stable response styles, Psychological Assessment 22(2), 366-381, https://doi.org/10.1037/a0019225; Mimura, C., & Griffiths, P. (2007), A Japanese version of the Rosenberg Self-Esteem Scale: Translation and equivalence assessment, Journal of Psychosomatic Research 62(5), 589-594, https://doi.org/10.1016/j.jpsychores.2006.11.004; Orth, U., & Robins, R. W. (2014), The development of self-esteem, Current Directions in Psychological Science 23(5), 381-387, https://doi.org/10.1177/0963721414547414; Oshio, A., Okada, R., Mogaki, M., Namikawa, T., & Wakita, T. (2014), Age and survey-year effects on self-esteem in Japan: A cross-temporal meta-analysis of scores on Rosenberg’s Self-Esteem Scale [in Japanese], Japanese Journal of Educational Psychology 62(4), 273-282, https://doi.org/10.5926/jjep.62.273; Robins, R. W., Hendin, H. M., & Trzesniewski, K. H. (2001), Measuring global self-esteem: Construct validation of a single-item measure and the Rosenberg Self-Esteem Scale, Personality and Social Psychology Bulletin 27(2), 151-161, https://doi.org/10.1177/0146167201272002; Rosenberg, M. (1965), Society and the adolescent self-image, Princeton University Press; Schmitt, D. P., & Allik, J. (2005), Simultaneous administration of the Rosenberg Self-Esteem Scale in 53 nations: Exploring the universal and culture-specific features of global self-esteem, Journal of Personality and Social Psychology 89(4), 623-642, https://doi.org/10.1037/0022-3514.89.4.623; Tomás, J. M., & Oliver, A. (1999), Rosenberg’s self-esteem scale: Two factors or method effects, Structural Equation Modeling 6(1), 84-98, https://doi.org/10.1080/10705519909540120; Trzesniewski, K. H., Donnellan, M. B., & Robins, R. W. (2003), Stability of self-esteem across the life span, Journal of Personality and Social Psychology 84(1), 205-220, https://doi.org/10.1037/0022-3514.84.1.205; Uchida, T., & Ueno, T. (2010), Reliability and validity of the Rosenberg Self Esteem Scale: Using the Japanese version of the RSES by Mimura & Griffiths (2007) [in Japanese], Annual Bulletin, Graduate School of Education, Tohoku University 58(2), 257-266, https://tohoku.repo.nii.ac.jp/record/3638/files/1346-5740-2010-58%282%29-257.pdf; University of Maryland, Department of Sociology (n.d.-a), Using the Rosenberg Self-Esteem Scale, https://socy.umd.edu/about-us/using-rosenberg-self-esteem-scale; University of Maryland, Department of Sociology (n.d.-b), Rosenberg Scale FAQ, https://socy.umd.edu/quick-links/rosenberg-scale-faq-0; Vallières, E. F., & Vallerand, R. J. (1990), Traduction et validation canadienne-française de l’échelle de l’estime de soi de Rosenberg, International Journal of Psychology 25(2), 305-316, https://doi.org/10.1080/00207599008247865; Yamamoto, M., Matsui, Y., & Yamanari, Y. (1982), Structure of perceived aspects of self [in Japanese], Japanese Journal of Educational Psychology 30(1), 64-68, https://doi.org/10.5926/jjep1953.30.1_64