Shuichiro Ogawa
日本語

Notes · updated 2026-11-01 · Markdown

Where Has Research on Non-Cognitive Skills Moved? Trends in Naming, Wage Returns, Interventions, and Measurement (2015–2026)

This note reviews what research on non-cognitive skills (social and emotional skills) has asked since 2015 and where its focus has moved. It draws on 47 sources, mostly peer-reviewed papers, together with counts of how often the terms appear.

Contents (13)
  1. How has the use of the term “non-cognitive” changed in English and Japanese?
  2. What was wrong with calling these qualities “not cognitive”?
  3. How have personality traits and skills been distinguished?
  4. Does the labor market pay more for social skills than before?
  5. How large have the effects of school-based SEL been estimated to be?
  6. What did randomized experiments show to persist?
  7. Do teachers and schools change something other than test scores?
  8. What becomes inverted when schools are compared through self-reports?
  9. Which skills decline with age?
  10. What has research in Japan asked?
  11. How should the trends of the decade be read?
  12. What this review can and cannot say
  13. Footnotes

How has the use of the term “non-cognitive” changed in English and Japanese?

In 2017, 229 English-language papers contained “non-cognitive skills” (or “noncognitive skills”) in the title or abstract. In 2025, the number was 233, almost the same. During the same period, the number of papers containing “skills” grew from 108,673 to 259,750. Per 10,000 papers containing “skills”, the term “non-cognitive” fell from 21.1 to 9.0 papers, while “social and emotional skills” (including “socio-emotional”) rose from 25.9 to 56.4 papers (counted in OpenAlex; the counting method is described in source/review/noncognitive-skills-trends/papers.md)1. Papers using the social and emotional terms were 1.23 times as many as those using “non-cognitive” in 2017 and 6.28 times as many in 2025.

Japanese-language papers moved in the opposite direction. In CiNii Research, the number of papers containing “hi-ninchi noryoku” (non-cognitive abilities) was 6 in 2014, 19 in 2017, and 108 in 2025. The OECD’s Japanese term for social and emotional skills, “shakai jodoteki sukiru”, appeared in 37 papers in 2025, about one third of the count for “hi-ninchi noryoku”. In other words, Japanese-language papers began to use the term “non-cognitive” during the same period in which English-language papers moved away from it.

The counts of terms do not show whether the content of research changed. To examine whether only the names changed or whether the way research asks its questions also moved, this note read 47 peer-reviewed papers and public reports, mainly from 2015 to 2026, and grouped them by question into nine clusters. Numbers and claims were checked against the full text for the 33 sources whose full text was available (the journal version, an author version, a preprint, or a working paper) and against the abstract for the remaining 14 (see the unverified items at the end). Heritability, grit, and growth mindset interventions are treated in Is the Ability to Make an Effort Also a Talent? The Heritability of Grit, Self-Control, and Motivation to Learn, and How Far These Traits Can Be Changed, and the general measurement of dispositions is treated in When Is a Disposition Being Measured? Measurement as Conditional Tendency, and Measurement as Co-occurrence; these topics are therefore not the focus here.

What was wrong with calling these qualities “not cognitive”?

Non-cognitive skills is a term that groups qualities outside the abilities measured by intelligence and achievement tests, such as perseverance, self-control, cooperation, and emotional stability. Duckworth and Yeager (2015) trace the origin of the term to Messick (1979). According to this account, once the term “cognitive” was assigned to intellectual abilities and school achievement, “noncognitive” came to the fore by default to describe everything else. Duckworth and Yeager regarded it as a problem that the term implies that some mental functioning involves no cognition. Their reason is that even self-control, a typical “non-cognitive” quality, depends on how temptations are represented in the mind.

The problem of naming was also a problem of measurement. Humphries and Kosse (2017) compared, within one data set, the ways of measuring non-cognitive skills used in labor, education, and behavioral economics. How the non-cognitive factor was constructed greatly changed both what the factor actually measured (personality traits or economic preferences) and the conclusions reached about educational attainment (abstract of the working paper version). Schoon (2021) noted that research on social and emotional competences has been afflicted by the jingle fallacy (the same term for different things) and the jangle fallacy (different terms for the same thing). In this sense, a single category of “non-cognitive” contained different qualities measured with different instruments.

Similar points have been made in Japan. Oshio (2023) noted that in Japan a set of psychological characteristics similar to non-cognitive skills and social and emotional skills is often called “hi-ninchi noryoku” (non-cognitive abilities). He discussed its definition by separating the conceptual setting, in which the term includes various characteristics other than cognitive abilities, from three conditions: whether the characteristics can be measured, whether they predict outcomes, and whether they can be changed by intervention. Muto et al. (2024) stated that, as the term “non-cognitive” spread to families, childcare, and schools, the idea that non-cognitive skills do not involve cognition may also have spread, and they examined what lies “in between” the cognitive and the non-cognitive.

How have personality traits and skills been distinguished?

During the same period, one framework that reorganized qualities other than cognition was the framework of the OECD Survey on Social and Emotional Skills (SSES). For this international survey of 10- and 15-year-olds, Chernyshenko et al. (2018) reviewed skills within and outside the five-factor (Big Five) model of personality by their structure, development, malleability, cross-cultural comparability, and relation to outcomes (abstract). Muto et al. (2024) point out that this framework largely follows the Big Five classification, so that traits, which have been assumed to be stable and hard to change, and skills or abilities, which can be raised through practice and learning, are treated on the same level.

The framework of SEB skills (social, emotional, and behavioral skills) proposed by Soto, Napolitano, and colleagues distinguished traits from skills explicitly. Napolitano et al. (2021) and Soto et al. (2022) defined skills as how someone is capable of thinking, feeling, and behaving when the situation calls for it, and traits as how someone tends to think, feel, and behave averaged across situations. Using five studies with seven samples (N = 6,309), Soto et al. (2022) developed the BESSI, an inventory of 32 skill facets in five domains: social engagement, cooperation, self-management, emotional resilience, and innovation. BESSI items ask respondents to rate how well they can perform a behavior on a five-point scale from “not at all well” to “extremely well”.

Whether the distinction matters was examined through prediction. In 897 high school students, the five BESSI domains significantly added explanatory power beyond the Big Five for 25 of 30 outcomes and competencies, with a mean increment of ΔR² = .06 (Soto et al. 2024, author version). In contrast, the Big Five added explanatory power beyond the BESSI for 13 of them, with a mean increment of .03. However, both skills and traits were measured by self-report, and the difference between them lies in whether the item asks what people can do or what they tend to do. Whether the two separate as different constructs across measurement methods was not examined in the studies read for this note.

Does the labor market pay more for social skills than before?

One reference point in economics is a study showing that a small number of dimensions of cognitive and noncognitive abilities explain schooling, wages, employment, and risky behavior (Heckman et al. 2006). Using the noncognitive ability that psychologists assessed in personal interviews at the Swedish military enlistment, men who fared poorly in the labor market, in the sense of unemployment or low annual earnings, lacked noncognitive rather than cognitive ability (Lindqvist and Vestman 2011, abstract). In the same study, cognitive ability was a stronger predictor of wages for skilled workers and of earnings above the median.

Research since 2015 has asked whether this value has changed over time. In the United States, jobs requiring high levels of social interaction grew by nearly 12 percentage points as a share of employment between 1980 and 2012, while jobs that are math-intensive but less social shrank by 3.3 percentage points (Deming 2017). The wage return to social skills was greater in the 2000s than in the 1980s and 1990s. In Swedish administrative data, the wage return to teamwork and leadership skills assessed by psychologists at conscription roughly doubled between 1992 and 2013, while the return to cognitive skill was relatively stable and decreased modestly during the 2000s (Edin et al. 2022). Among about half a million Finnish men (2001–2015), the value of cognitive skills declined, while the value of noncognitive skills increased (Izadi and Tuhkuri 2026). This rise was driven by extraversion, and the value of conscientiousness remained stable.

In Finland, therefore, the value rose for only some of the “non-cognitive” qualities. How social skills work within groups was examined in an experiment by Weidmann and Deming (2021). They randomly assigned people to multiple teams and identified as “team players” those who consistently caused their team to exceed the performance predicted from the members’ previously assessed skills. Team players scored higher on a test of social intelligence but did not differ in IQ, personality, education, or gender. Social skills improved group performance about as much as IQ (working paper version).

The association between personality and life outcomes itself was also subjected to replication. In preregistered replications of 78 associations between the Big Five and life outcomes (median N = 1,504), 87% were statistically significant in the expected direction, and the effects were about 77% as strong as the original effects (Soto 2019, abstract). Most associations replicated, but they were smaller than originally reported.

Whether demand for social skills will increase in the era of generative AI is not yet settled. The latest version of a preprint that analyzed about seven million job postings from 595 U.S. public firms adopting generative AI, using difference-in-differences models, reported that the relative demand for social skills in roles adopting generative AI declined by 3.4%, while demand for cognitive skills held steady (Gulati et al. 2025). The first version of the same preprint (March 2025, 378 firms) had reported that demand for social skills increased by 5.2%. The preprint has not been peer reviewed, and the direction of its conclusion reversed between versions.

How large have the effects of school-based SEL been estimated to be?

Programs that teach social and emotional skills in school lessons are called social and emotional learning (SEL). Their effects have been estimated in meta-analyses of universal school-based SEL. Durlak et al. (2011) reported, for 213 programs and 270,034 students from kindergarten through high school, that participants outperformed controls in social and emotional skills, attitudes, behavior, and academic performance, with the improvement in achievement corresponding to 11 percentile points (abstract). According to Cipriano et al. (2023), the effect sizes in Durlak et al. were g = 0.57 for social and emotional skills and g = 0.27 for academic achievement. Taylor et al. (2017) reported, for 82 interventions and 97,406 students followed up from 6 months to 18 years after the intervention, that participants still fared better than controls at follow-up in social and emotional skills, attitudes, and well-being (abstract).

In a meta-analysis conducted more than ten years later, the effects were estimated to be smaller. Cipriano et al. (2023) synthesized 424 studies from 53 countries published from 2008 through 2020 (252 interventions, 575,361 students) using a three-level model. In the full text of the version under review, the overall effect was g = 0.194, the effect on SEL skills was g = 0.219, the effect on academic achievement was g = 0.109, and the effect at follow-up was g = 0.1672. The effects differed considerably by SEL content, intervention features, context, and implementation quality (abstract of the journal version). Because the 2011 and 2023 meta-analyses differ in the periods of the studies they collected and in their estimation methods, the whole difference in effect sizes cannot be read as evidence that the effects shrank. However, other meta-analyses have also discussed how effects differ with the rigor of studies and the conditions of evaluation. A meta-analysis focusing on academic achievement (40 studies) reported effects of +0.25 for reading, +0.26 for mathematics, and +0.19 for science, while stating that SEL programs examined in more rigorous randomized studies with large samples might not have as meaningful effects as once thought (Corcoran et al. 2018, abstract). A meta-analysis of 89 studies reported that effects differed by the stage of evaluation, the involvement of the program developer, and whether the program was implemented outside its country of origin, and that these differences were not always in the hypothesized direction (Wigelsworth et al. 2016, abstract).

For research in Japan, the quality of evidence is the first concern. In a meta-analysis of 85 studies with control groups involving Japanese children aged 5 to 18, the overall effect was ES = 0.26 (Takizawa et al. 2023). None of the 85 studies used a randomized design that was judged to have a low risk of bias.

Personality traits themselves have also been reported to change through intervention. In a meta-analysis of 207 studies, mainly of clinical interventions, personality trait measures changed by d = .37 over an average of 24 weeks of intervention, and emotional stability changed the most (Roberts et al. 2017, abstract). Most of these changes were observed in people receiving psychotherapy, not in school students.

What did randomized experiments show to persist?

In economics, studies have appeared that randomize a small number of interventions and follow them for a long time. A starting point was the reanalysis of the Perry Preschool Program in the United States. The program did not produce long-run gains in IQ, but it produced persistent improvements in personality skills such as externalizing behavior (aggressive, antisocial, and rule-breaking behavior), and these improvements explained a sizable portion of the treatment effects in adulthood (Heckman et al. 2013).

The experiments since 2018 read for this note randomly assigned curricula or mentors that develop specific qualities to primary school children. When teachers taught a curriculum that encouraged children to imagine their future selves, the children made more patient choices in incentivized tasks, and the effect persisted almost three years after the intervention and replicated in a different sample (Alan and Ertac 2018, abstract). A grit curriculum that taught the plasticity of the brain, the role of effort, the constructive interpretation of failure, and goal setting left an effect of about 0.2 standard deviations on a standardized mathematics test 2.5 years later in two independent primary school samples (Alan et al. 2019). When primary school children in Germany were given a mentor for one year, their prosociality increased, and the increase persisted at a two-year follow-up (Kosse et al. 2020, working paper version). The authors state that the program may close the gap in prosociality by socioeconomic status.

The longest follow-up concerns an intervention that taught social and emotional skills in primary school classrooms in Switzerland. The effects persisted for more than a decade, and treated children were more likely to complete academic high school and to enroll in university (Sorrenti et al. 2025). As mechanisms, symptoms of attention deficit hyperactivity disorder (impulsivity and disruptive behavior) decreased and grades improved, but scores on standardized tests did not increase. The authors interpret the long-term effects on educational attainment as driven by changes in social and emotional skills rather than cognitive skills. In the Perry Preschool Program and in this intervention, the remaining effects appeared in behavior, grades, and educational attainment rather than in test scores. In contrast, the effect of the grit curriculum (Alan et al. 2019) appeared in a standardized mathematics test.

Do teachers and schools change something other than test scores?

Teacher quality has long been measured by gains in students’ test scores (value-added). Jackson (2018) constructed a behavior index from ninth-grade absences, suspensions, grades, and grade repetition, and estimated teacher effects on this index. Teacher effects on behavior were only weakly correlated with teacher effects on test scores (r = 0.22 in the working paper version), and effects on behavior predicted high school completion and longer-run outcomes more strongly than effects on test scores. Using both measures more than doubled the variance of predictable teacher effects on longer-run outcomes compared with using test scores alone.

Results of the same form were repeated in different settings and with different measures. In a study using the random assignment of class rosters, teacher effects on complex open-ended tasks in mathematics and reading, growth mindset, grit, and effort in class were large but only weakly related to value-added on state tests (Kraft 2019, abstract). Teacher effects on reducing unexcused absences were only weakly correlated with effects on achievement and had a stronger effect on high school completion (Liu and Loeb 2021, submitted version). In a simulation that replaced teachers using both test and non-test effects, most long-term outcomes improved by more than 50% compared with a policy using test scores alone (Petek and Pope 2023, abstract). High schools in Chicago improved students’ self-reported social and emotional development, and this school value-added predicted longer-run outcomes better than test-score value-added (Jackson et al. 2020).

What becomes inverted when schools are compared through self-reports?

If teachers and schools have effects outside test scores, instruments to measure them are needed. At this point, the weaknesses of self-report questionnaires became visible. In 1,368 eighth graders in the United States, self-reports of conscientiousness, self-control, grit, and growth mindset were related to attendance, behavior, and test-score gains at the student level (West et al. 2016, abstract). At the school level, however, conscientiousness, self-control, and grit were unrelated to test-score gains. Oversubscribed charter schools, admission to which was decided by lottery, raised achievement and attendance but lowered scores on these self-reported measures. The authors suggested that reference bias, the tendency for survey responses to be influenced by the surrounding standards, may have produced this paradox.

Lira et al. (2022) showed, in three studies of 229,685 adolescents, that the higher the achievement of their peers, the lower the students rated their own self-regulation and the higher the standards they held. This pattern was not observed in task measures of self-regulation. Because of reference bias, self-reported self-regulation produced paradoxical predictions of college persistence six years later. On the other hand, the Chicago study stated that high school effects on social and emotional development could be captured by self-report surveys (Jackson et al. 2020). These results can be read as indicating that whether self-reports can be used to compare schools depends on how much the standards differ across schools.

Alternative measures have also been tested. A study of 513 high school seniors showed the potential of item nonresponse and careless answering in surveys as proxy measures of grit and self-control, based on their relation to teacher reports, performance tasks, high school outcomes, and college attendance (Zamarro et al. 2020). Measures of student effort constructed from the PISA test and student survey explained 32 to 38 percent of the variation in test scores across countries (Zamarro et al. 2019, abstract). When the decline in performance during the PISA test was treated as a measure of noncognitive skills, it was positively associated with economic growth together with the starting level of performance (Balart et al. 2018, working paper version). In some cases, the questionnaire itself does not work across cultures. In 29 face-to-face surveys of 94,751 respondents in 23 low- and middle-income countries, commonly used personality questions failed to measure the intended traits, whereas the validity was much higher among 198,356 internet respondents from the same countries (Laajaj et al. 2019).

Which skills decline with age?

On average, personality traits change with age in the direction of maturity. A meta-analysis of longitudinal studies published after 2005 covered 189 samples and 178,503 people for rank-order stability and 276 samples and 242,542 people for mean-level change (Bleidorn et al. 2022). Rank-order stability increased from childhood to young adulthood and then reached a plateau, cumulative mean-level change across the life span was slightly smaller than previously estimated, and emotional stability increased more consistently and substantially than previously found.

Age differences in skills showed domains that decline in adolescence. In a cross-sectional study of 4,106 people aged 12 to 19 (a sample collected by the authors in Italy combined with open data from Germany and the United States), emotional resilience and cooperation skills increased with age, while innovation, social engagement, and self-management skills declined, especially between 12 and 16 years, and grew later (Feraco and Meneghetti 2023). Because the comparison is cross-sectional, it does not show that the same individuals declined. The OECD SSES 2023 is the largest international survey of these skills among 10- and 15-year-olds and covered 16 countries and cities, including Gunma Prefecture in Japan. It reported that social and emotional skills are related to academic achievement, life satisfaction, healthier behavior, lower anxiety, and more ambitious career plans, and that they are inequitably distributed by age, gender, and socioeconomic background (OECD 2024, abstract). The report also compares the results with the 2019 survey, conducted before the pandemic. In two measurements of 512 pairs of adolescents and parents in three European countries during the pandemic (October 2020 and May 2021), social and emotional skills tended to decline, but the changes were not conspicuous, and most were not statistically significant (Martinsone et al. 2022).

What has research in Japan asked?

On when the term “non-cognitive” spread in Japan, Muto et al. (2024) state that books and translations related to the non-cognitive appeared from around 2015, and that James J. Heckman and the OECD played a large role in its spread. A project of the National Institute for Educational Policy Research (NIER) in fiscal years 2015 and 2016 (Endo et al. 2017) stated that what non-cognitive abilities are had not been sufficiently clarified; it therefore focused on social-emotional competence, reviewed prior research, and measured children directly to examine the situation in Japan.

Subsequent empirical research in Japan has followed two lines: reanalyses of existing large-scale data and estimates of intervention effects. Yoshizawa et al. (2026) conducted a factor analysis of the student questionnaire in the census data of sixth graders from the fiscal year 2015 National Assessment of Academic Ability and extracted ten factors. They found significant associations between subscales corresponding to non-cognitive abilities, such as “engagement with written-answer questions”, “interest, motivation, and attitude”, and “basic daily habits”, and subject test scores. On the intervention side, as noted above, none of the 85 studies used a randomized design with a low risk of bias (Takizawa et al. 2023). Within the sources read for this note, research in Japan still includes few designs that estimate causal effects.

When the nine clusters are placed side by side, research, within the sources read for this note, has moved in three directions behind the change of terms.

The first is a move away from grouping everything that is not cognitive under one term. The term “non-cognitive” arose as a default name (Duckworth and Yeager 2015), conclusions changed with the construction of the factor (Humphries and Kosse 2017), and the field carried the jingle and jangle fallacies (Schoon 2021). English-language research separated personality traits from skills (Soto et al. 2022) and shifted to the term social and emotional skills. However, the attempt to separate them still rests on differences in how the same self-report questions are asked.

The second is a move from correlations of value to the causal effects of interventions, and further to outcomes outside test scores. Research on wage returns showed, with administrative data from several countries, that the value of social skills rose over time (Deming 2017; Edin et al. 2022; Izadi and Tuhkuri 2026). Intervention research showed, through randomization and long follow-up, examples in which effects remained in behavior, grades, and educational attainment rather than in test scores (Heckman et al. 2013; Sorrenti et al. 2025). Research on teachers and schools extracted effects that are only weakly correlated with effects on test scores, using absences, behavior, and self-reports (Jackson 2018; Petek and Pope 2023; Jackson et al. 2020). During the same period, meta-analytic estimates of the effects of school-based SEL became smaller (the comparison between Durlak et al. 2011 and Cipriano et al. 2023; Corcoran et al. 2018).

The third is doubt about comparisons based on self-report. Reference bias can invert comparisons between schools (West et al. 2016; Lira et al. 2022). For this reason, measures based on behavioral traces, such as effort in answering and the decline in performance during a test, have increased (Zamarro et al. 2019, 2020; Balart et al. 2018). The more the second direction requires instruments for outcomes outside test scores, the heavier the weaknesses of measurement shown by the third direction become.

Research in Japan has taken up the debate of the first direction (Oshio 2023; Muto et al. 2024) and has begun to reanalyze the questionnaire of a national survey (Yoshizawa et al. 2026), but, within the sources read for this note, causal designs and examinations of bias in school comparisons based on self-report remain few. Whether generative AI raises or lowers the value of social skills remains an open question, as even the versions of one preprint disagree.

What this review can and cannot say

This review can say that, in English-language papers, the term “non-cognitive” decreased relative to the social and emotional terms, and that in Japanese-language papers the term “hi-ninchi noryoku” increased from around 2017. It can also say that the studies read for this note concentrate on three issues: the distinction between personality traits and skills, the causal links to wages and long-term outcomes, and bias in comparisons based on self-report.

There is also much that it cannot say. The 47 sources were selected for each issue and are not a sample of the whole body of research on non-cognitive skills. The reading of three directions is an interpretation based on repetition within the selected studies, not a measurement based on publication or citation counts. The counts of terms depend on how OpenAlex and CiNii Research record papers, and in OpenAlex the records with abstracts and non-English records increased after 2024. The effect sizes of Cipriano et al. (2023) are from the full text of the version under review, and those of Durlak et al. (2011) are as cited by Cipriano et al. Research on wage returns relies mainly on a U.S. national survey and on Nordic data from the conscription and administrative records of men, and its applicability to Japan and to Nordic women was not examined. The randomized experiments each concern one curriculum or mentoring program for primary school children in one country (Turkey in Alan et al. 2019, Germany in Kosse et al. 2020, and Switzerland in Sorrenti et al. 2025), and their generalization to other countries and programs has not been examined. Only the abstract of the OECD SSES report could be read, so the size of the differences between 10- and 15-year-olds is not reported here. The relation between generative AI and social skills is based on one preprint that has not been peer reviewed.

Unverified items

No unverified statements remain in the main claims. The full text of the following 14 sources could not be obtained (the publisher pages refused automated access, and no open-access version was found), so they were described within the scope of the abstract and marked as such in the text. Chernyshenko et al. (2018), Lindqvist and Vestman (2011), Soto (2019), Durlak et al. (2011), Taylor et al. (2017), Corcoran et al. (2018), Wigelsworth et al. (2016), Roberts et al. (2017), Alan and Ertac (2018), Kraft (2019), Petek and Pope (2023), West et al. (2016), Zamarro et al. (2019), and OECD (2024). For sources whose journal full text could not be obtained and which were checked against a working paper, a preprint, or an author version, the numbers were compared with the abstract of the journal version. Numbers not in the abstract (r = 0.22 in Jackson 2018, ΔR² in Soto et al. 2024, and the effect sizes in Cipriano et al. 2023) are attributed to the version in the text. Gulati et al. (2025) is a preprint that has not been peer reviewed, and no peer-reviewed version was confirmed.

References

All sources were accessed on 2026-11-01. The ledger and verification records are in source/review/noncognitive-skills-trends/papers.md and fulltext-manifest.json.

Counts of terms

Naming and frameworks

  • Chernyshenko, Oleksandr S., Miloš Kankaraš, and Fritz Drasgow. 2018. “Social and Emotional Skills for Student Success and Well-Being: Conceptual Framework for the OECD Study on Social and Emotional Skills.” OECD Education Working Papers. OECD Publishing. https://doi.org/10.1787/db1d8e59-en
  • Duckworth, Angela L., and David Scott Yeager. 2015. “Measurement Matters: Assessing Personal Qualities Other Than Cognitive Ability for Educational Purposes.” Educational Researcher 44 (4): 237–251. https://doi.org/10.3102/0013189X15584327
  • Heckman, James J., and Tim Kautz. 2012. “Hard Evidence on Soft Skills.” Labour Economics 19 (4): 451–464. https://doi.org/10.1016/j.labeco.2012.05.014
  • Humphries, John Eric, and Fabian Kosse. 2017. “On the Interpretation of Non-Cognitive Skills: What Is Being Measured and Why It Matters.” Journal of Economic Behavior & Organization 136: 174–185. https://doi.org/10.1016/j.jebo.2017.02.001
  • Napolitano, Christopher M., Madison N. Sewell, Hee J. Yoon, Christopher J. Soto, and Brent W. Roberts. 2021. “Social, Emotional, and Behavioral Skills: An Integrative Model of the Skills Associated With Success During Adolescence and Across the Life Span.” Frontiers in Education 6: 679561. https://doi.org/10.3389/feduc.2021.679561
  • Schoon, Ingrid. 2021. “Towards an Integrative Taxonomy of Social-Emotional Competences.” Frontiers in Psychology 12: 515313. https://doi.org/10.3389/fpsyg.2021.515313
  • Soto, Christopher J., Christopher M. Napolitano, Madison N. Sewell, Hee J. Yoon, and Brent W. Roberts. 2022. “An Integrative Framework for Conceptualizing and Assessing Social, Emotional, and Behavioral Skills: The BESSI.” Journal of Personality and Social Psychology 123 (1): 192–222. https://doi.org/10.1037/pspp0000401
  • Soto, Christopher J., Christopher M. Napolitano, Madison N. Sewell, Hee Jun Yoon, and Brent W. Roberts. 2024. “Going Beyond Traits: Social, Emotional, and Behavioral Skills Matter for Adolescents’ Success.” Social Psychological and Personality Science 15 (1): 33–45. https://doi.org/10.1177/19485506221127483

Labor market and life outcomes

  • Deming, David J. 2017. “The Growing Importance of Social Skills in the Labor Market.” Quarterly Journal of Economics 132 (4): 1593–1640. https://doi.org/10.1093/qje/qjx022
  • Edin, Per-Anders, Peter Fredriksson, Martin Nybom, and Björn Öckert. 2022. “The Rising Return to Noncognitive Skill.” American Economic Journal: Applied Economics 14 (2): 78–100. https://doi.org/10.1257/app.20190199
  • Gulati, Piyush, Arianna Marchetti, Victoria Sevcenko, and Phanish Puranam. 2025. “How Generative AI Adoption Alters the Demand for Cognitive and Social Skills Within Roles: A Skill-Centric Analysis.” arXiv:2503.09212 (preprint). https://arxiv.org/abs/2503.09212
  • Heckman, James J., Jora Stixrud, and Sergio Urzua. 2006. “The Effects of Cognitive and Noncognitive Abilities on Labor Market Outcomes and Social Behavior.” Journal of Labor Economics 24 (3): 411–482. https://doi.org/10.1086/504455
  • Izadi, Ramin, and Joonas Tuhkuri. 2026. “Beyond IQ: The Rising Value of Extraversion.” Journal of Labor Economics 44 (3): 925–966. https://doi.org/10.1086/733226
  • Lindqvist, Erik, and Roine Vestman. 2011. “The Labor Market Returns to Cognitive and Noncognitive Ability: Evidence from the Swedish Enlistment.” American Economic Journal: Applied Economics 3 (1): 101–128. https://doi.org/10.1257/app.3.1.101
  • Soto, Christopher J. 2019. “How Replicable Are Links Between Personality Traits and Consequential Life Outcomes? The Life Outcomes of Personality Replication Project.” Psychological Science 30 (5): 711–727. https://doi.org/10.1177/0956797619831612
  • Weidmann, Ben, and David J. Deming. 2021. “Team Players: How Social Skills Improve Team Performance.” Econometrica 89 (6): 2637–2657. https://doi.org/10.3982/ECTA18461

School-based SEL and personality interventions

  • Cipriano, Christina, Michael J. Strambler, Lauren H. Naples, Cheyeon Ha, Megan Kirk, Miranda Wood, et al. 2023. “The State of Evidence for Social and Emotional Learning: A Contemporary Meta-Analysis of Universal School-Based SEL Interventions.” Child Development 94 (5): 1181–1204. https://doi.org/10.1111/cdev.13968
  • Corcoran, Roisin P., Alan C. K. Cheung, Elizabeth Kim, and Chen Xie. 2018. “Effective Universal School-Based Social and Emotional Learning Programs for Improving Academic Achievement: A Systematic Review and Meta-Analysis of 50 Years of Research.” Educational Research Review 25: 56–72. https://doi.org/10.1016/j.edurev.2017.12.001
  • Durlak, Joseph A., Roger P. Weissberg, Allison B. Dymnicki, Rebecca D. Taylor, and Kriston B. Schellinger. 2011. “The Impact of Enhancing Students’ Social and Emotional Learning: A Meta-Analysis of School-Based Universal Interventions.” Child Development 82 (1): 405–432. https://doi.org/10.1111/j.1467-8624.2010.01564.x
  • Roberts, Brent W., Jing Luo, Daniel A. Briley, Philip I. Chow, Rong Su, and Patrick L. Hill. 2017. “A Systematic Review of Personality Trait Change Through Intervention.” Psychological Bulletin 143 (2): 117–141. https://doi.org/10.1037/bul0000088
  • Takizawa, Yu, Matthew Bambling, Yuki Matsumoto, Yuma Ishimoto, and Sisira Edirippulige. 2023. “Effectiveness of Universal School-Based Social-Emotional Learning Programs in Promoting Social-Emotional Skills, Attitudes towards Self and Others, Positive Social Behaviors, and Improving Emotional and Conduct Problems among Japanese Children: A Meta-Analytic Review.” Frontiers in Education 8: 1228269. https://doi.org/10.3389/feduc.2023.1228269
  • Taylor, Rebecca D., Eva Oberle, Joseph A. Durlak, and Roger P. Weissberg. 2017. “Promoting Positive Youth Development Through School-Based Social and Emotional Learning Interventions: A Meta-Analysis of Follow-Up Effects.” Child Development 88 (4): 1156–1171. https://doi.org/10.1111/cdev.12864
  • Wigelsworth, M., A. Lendrum, J. Oldfield, A. Scott, I. ten Bokkel, K. Tate, et al. 2016. “The Impact of Trial Stage, Developer Involvement and International Transferability on Universal Social and Emotional Learning Programme Outcomes: A Meta-Analysis.” Cambridge Journal of Education 46 (3): 347–376. https://doi.org/10.1080/0305764X.2016.1195791

Randomized experiments and long-term follow-up

  • Alan, Sule, Teodora Boneva, and Seda Ertac. 2019. “Ever Failed, Try Again, Succeed Better: Results from a Randomized Educational Intervention on Grit.” Quarterly Journal of Economics 134 (3): 1121–1162. https://doi.org/10.1093/qje/qjz006
  • Alan, Sule, and Seda Ertac. 2018. “Fostering Patience in the Classroom: Results from Randomized Educational Intervention.” Journal of Political Economy 126 (5): 1865–1911. https://doi.org/10.1086/699007
  • Heckman, James, Rodrigo Pinto, and Peter Savelyev. 2013. “Understanding the Mechanisms Through Which an Influential Early Childhood Program Boosted Adult Outcomes.” American Economic Review 103 (6): 2052–2086. https://doi.org/10.1257/aer.103.6.2052
  • Kosse, Fabian, Thomas Deckers, Pia Pinger, Hannah Schildberg-Hörisch, and Armin Falk. 2020. “The Formation of Prosociality: Causal Evidence on the Role of Social Environment.” Journal of Political Economy 128 (2): 434–467. https://doi.org/10.1086/704386
  • Sorrenti, Giuseppe, Ulf Zölitz, Denis Ribeaud, and Manuel Eisner. 2025. “The Causal Impact of Socio-Emotional Skills Training on Educational Success.” Review of Economic Studies 92 (1): 506–552. https://doi.org/10.1093/restud/rdae018

Effects of teachers and schools

  • Jackson, C. Kirabo. 2018. “What Do Test Scores Miss? The Importance of Teacher Effects on Non–Test Score Outcomes.” Journal of Political Economy 126 (5): 2072–2107. https://doi.org/10.1086/699018
  • Jackson, C. Kirabo, Shanette C. Porter, John Q. Easton, Alyssa Blanchard, and Sebastián Kiguel. 2020. “School Effects on Socioemotional Development, School-Based Arrests, and Educational Attainment.” American Economic Review: Insights 2 (4): 491–508. https://doi.org/10.1257/aeri.20200029
  • Kraft, Matthew A. 2019. “Teacher Effects on Complex Cognitive Skills and Social-Emotional Competencies.” Journal of Human Resources 54 (1): 1–36. https://doi.org/10.3368/jhr.54.1.0916.8265R3
  • Liu, Jing, and Susanna Loeb. 2021. “Engaging Teachers: Measuring the Impact of Teachers on Student Attendance in Secondary School.” Journal of Human Resources 56 (2): 343–379. https://doi.org/10.3368/jhr.56.2.1216-8430R3
  • Petek, Nathan, and Nolan G. Pope. 2023. “The Multidimensional Impact of Teachers on Students.” Journal of Political Economy 131 (4): 1057–1107. https://doi.org/10.1086/722227

Measurement

  • Balart, Pau, Matthijs Oosterveen, and Dinand Webbink. 2018. “Test Scores, Noncognitive Skills and Economic Growth.” Economics of Education Review 63: 134–153. https://doi.org/10.1016/j.econedurev.2017.12.004
  • Laajaj, Rachid, Karen Macours, Daniel Alejandro Pinzon Hernandez, Omar Arias, Samuel D. Gosling, Jeff Potter, et al. 2019. “Challenges to Capture the Big Five Personality Traits in Non-WEIRD Populations.” Science Advances 5 (7): eaaw5226. https://doi.org/10.1126/sciadv.aaw5226
  • Lira, Benjamin, Joseph M. O’Brien, Pablo A. Peña, Brian M. Galla, Sidney D’Mello, David S. Yeager, et al. 2022. “Large Studies Reveal How Reference Bias Limits Policy Applications of Self-Report Measures.” Scientific Reports 12: 19189. https://doi.org/10.1038/s41598-022-23373-9
  • West, Martin R., Matthew A. Kraft, Amy S. Finn, Rebecca E. Martin, Angela L. Duckworth, Christopher F. O. Gabrieli, et al. 2016. “Promise and Paradox: Measuring Students’ Non-Cognitive Skills and the Impact of Schooling.” Educational Evaluation and Policy Analysis 38 (1): 148–170. https://doi.org/10.3102/0162373715597298
  • Zamarro, Gema, Collin Hitt, and Ildefonso Mendez. 2019. “When Students Don’t Care: Reexamining International Differences in Achievement and Student Effort.” Journal of Human Capital 13 (4): 519–552. https://doi.org/10.1086/705799
  • Zamarro, Gema, Malachi Nichols, Angela L. Duckworth, and Sidney K. D’Mello. 2020. “Validation of Survey Effort Measures of Grit and Self-Control in a Sample of High School Students.” PLOS ONE 15 (7): e0235396. https://doi.org/10.1371/journal.pone.0235396

Development and international comparison

  • Bleidorn, Wiebke, Ted Schwaba, Anqing Zheng, Christopher J. Hopwood, Susana S. Sosa, Brent W. Roberts, and D. A. Briley. 2022. “Personality Stability and Change: A Meta-Analysis of Longitudinal Studies.” Psychological Bulletin 148 (7–8): 588–619. https://doi.org/10.1037/bul0000365
  • Feraco, Tommaso, and Chiara Meneghetti. 2023. “Social, Emotional, and Behavioral Skills: Age and Gender Differences at 12 to 19 Years Old.” Journal of Intelligence 11 (6): 118. https://doi.org/10.3390/jintelligence11060118
  • Martinsone, Baiba, Ieva Stokenberga, Ilze Damberga, Inga Supe, Celeste Simões, Paula Lebre, et al. 2022. “Adolescent Social Emotional Skills, Resilience and Behavioral Problems during the COVID-19 Pandemic: A Longitudinal Study in Three European Countries.” Frontiers in Psychiatry 13: 942692. https://doi.org/10.3389/fpsyt.2022.942692
  • OECD. 2024. Social and Emotional Skills for Better Lives: Findings from the OECD Survey on Social and Emotional Skills 2023. OECD Publishing. https://doi.org/10.1787/35ca7b7c-en

Research in Japan

  • Endo, Toshihiko (principal investigator), et al. 2017. Report on Research on the Development of Non-Cognitive (Social-Emotional) Abilities and Scientific Methods for Examining Them (非認知的（社会情緒的）能力の発達と科学的検討手法についての研究に関する報告書). Project Research Report, Elementary and Secondary Education 031. National Institute for Educational Policy Research (in Japanese). https://www.nier.go.jp/05_kenkyu_seika/pdf_seika/h28a/syocyu-2-1_a.pdf
  • Muto, Sera, Ikuko Shinohara, Etsuko Haryu, Makiko Naka, and Machiko Sannomiya. 2024. “Cognitive, Noncognitive, and In-between: From the Perspectives of Education, Research, and Practice” (認知と非認知のあいだ：教育・研究・実践の観点から). The Annual Report of Educational Psychology in Japan 63: 207–225 (in Japanese). https://doi.org/10.5926/arepj.63.207
  • Oshio, Atsushi. 2023. “Some Issues Related to ‘Noncognitive Abilities’: From the Perspectives of Measurement, Prediction, and Intervention” (「非認知能力」の諸問題：測定・予測・介入の観点から). The Annual Report of Educational Psychology in Japan 62: 165–183 (in Japanese). https://doi.org/10.5926/arepj.62.165
  • Yoshizawa, Hiroyuki, Iori Tani, Hideharu Sugimoto, Asuka Matsumoto, Kanako Terao, and Tomoko Tamura. 2026. “A Comprehensive Reanalysis of the Questionnaires in the National Assessment of Academic Ability with Elementary School Data: Focusing on Noncognitive Skills” (全国学力・学習状況調査における児童質問紙調査の再分析：非認知能力に焦点化した検討). The Japanese Journal of Personality 34 (3): 299–314 (in Japanese). https://doi.org/10.2132/personality.34.3.8

Footnotes

  1. Counted by publication year with title_and_abstract.search in OpenAlex (2026-11-01). “Non-cognitive” is the OR of “non-cognitive skills”, “noncognitive skills”, and their singular forms; the social and emotional terms are the OR of “socio-emotional skills”, “socioemotional skills”, “social and emotional skills”, and “social-emotional skills”. The denominator, papers containing “skills”, was 108,673 in 2017 and 259,750 in 2025. SEL (“social and emotional learning” or “social-emotional learning”) appeared in 420 papers in 2017 and 1,914 papers in 2025. For CiNii Research, the count is the number of results of an article search across all fields. ↩

  2. The full text of the journal version was behind a paywall. The values are from the full text of the Stage 2 version under review (a Registered Report) posted on OSF. The abstract of the journal version is identical to the abstract of this version. ↩

Author: Shuichiro Ogawa (Design Researcher / Design Consultant) About me →