Because intellectually gifted individuals constitute a special population, and one that was not systematically studied, a great deal of information is still to be provided regarding related specificities. Many people have very different ideas about giftedness, including stereotypes or simply external perceptions that are not necessarily accurate, which can affect the way gifted youth are treated at school and in their everyday life (Barbier et al. 2022; Baudson 2016; Bergold et al. 2021; Carman 2011; Manaster et al. 1994; Subotnik et al. 2011; and Vialle et al. 2007).

Empirical studies support the conclusion that cognitive characteristics are associated with giftedness, such as high processing speed, effective representation of problems, flexibility in the choice of strategies and solutions, a broader knowledge base, etc. (Aubry et al. 2021; Calero et al. 2011; Geake 2008; Rodríguez Naveiras et al. 2019; and Steiner and Carr 2003). Nevertheless, it remains difficult to establish from the literature whether or not there are non-cognitive characteristics associated with giftedness. Indeed, despite the existence of a growing body of research interested in giftedness, no integrative literature review is available to synthesize findings about the non-cognitive specificities of gifted youth. However, public interest is largely focused on this non-cognitive profile of giftedness, which is often negatively connoted (Bergold et al. 2021). The implications of such a profile for clinical practice and education are very important. The daily lives, academic success, and well-being of gifted youth could be affected by these characteristics.

ReThisearchers aim article aims to fill this gap in the literature while confronting misconceptions about giftedness with available empirical findings. ResearchersTo prepare this review, we selected major topics regarding specificities associated with giftedness that come up frequently in discussions with laypeople and in media representations of giftedness. ResearchersWe only considered non-cognitive specificities in the sense that the topics covered here are not primarily related to cognitive ability as measured by a performance test; of course, this does not mean that the various topics discussed in this review are totally devoid of cognition (for example, emotional intelligence and humor are partly about cognitive processing).

Because there is no universal criterion for identification (McBee and Makel 2019), it is important to be aware that the groups of gifted individuals identified in studies in the literature do not necessarily refer to a unique subpopulation of individuals (Carman 2013). In fact, the literature distinguishes several types of giftedness or even talent (Olszewski-Kubilius et al. 2016; Sternberg and Davidson 2005). Multidimensional models of giftedness and talent were proposed (e.g., Gagné 2005; Renzulli 2005). These models include cognitive variables, such as high intellectual ability or academic excellence, as well as non-cognitive variables, such as leadership, motivation, or talent. The major problem with these models is the lack of information on the specific functional relationships between their components, which hampers their use in empirical research and the interpretation of the resulting findings (Wirthwein et al. 2019; for further discussion of the implications of these different approaches for clinical practice, see McBee and Makel 2019). For this reason, rwesearchers chose to focus the is systematic review of the literature on a single component of giftedness: intellectual giftedness defined as high intellectual ability. This one-dimensional approach to giftedness is not unusual; in fact, it was the most prevalent approach in the literature for a long time (Carman 2013) and is still widely used today. Some authors may consider it a restrictive approach to giftedness. ResWearchers believe that the criteria for including gifted groups should be different depending on whether researcherswe are talking about intellectual giftedness, creative giftedness, talent, or leadership. This distinction is essential from theour point of view until researcherswe have more theoretical knowledge of how the different elements that make up multidimensional models of giftedness relate to each other, particularly because what applies to intellectual giftedness is not necessarily transposable to other types of giftedness.

In many cases, more research was conducted regarding the correlation between a particular specificity and general intelligence than regarding the specific case of giftedness.

2. Methodological Issues

2.1. Inclusion Criteria

A major limitation of existing literature on giftedness is the wide heterogeneity in definitions and inclusion criteria for giftedness, which makes it particularly difficult to compare studies. ReIn thisearchers  literature review, we focused only on studies based on a definition of giftedness as high intellectual ability. ResearchersWe made this choice because researcherswe believe it is critical for inclusion criteria in gifted groups to be distinguished depending on whether one is talking about intellectual giftedness, creative giftedness, talent, or leadership, since what applies to intellectual giftedness does not necessarily apply to other types of giftedness. Even when focusing on intellectual giftedness, the articles selected for this literature review used different inclusion criteria for gifted groups, as can be seen in the summary tables. Some issues are important to discuss. First, most studies reviewed here relied solely or partly on achievement test scores and/or academic achievement (Carman 2013), presuming that high academic scores necessarily reflect intellectual giftedness. Although intelligence level does correlate with academic achievement, it is misleading to systematically associate potential with achievement. This approach fails to detect intellectually gifted children and adolescents who are under-achieving, either as a coping strategy (Winner 2000) or as a consequence of disengagement from school. Conversely, the socio-cultural context of a child (e.g., income of parents) strongly impacts their academic results, which is a major limitation of the use of achievement-based criteria regarding intellectual giftedness (Hanushek et al. 2019). Such an inclusion criterion leads to a truncated representation of the gifted samples by focusing on a part of the population that may have particular profiles with regard to numerous variables, such as, for example, the relationship between leadership and achievement. Another inclusion criterion sometimes applied to gifted groups is teacher nomination. This method considers that teachers are able to identify intellectually gifted students within their classes based on a range of academic as well as psychological and socio-emotional criteria. However, the results from the present literature review show that non-cognitive characteristics commonly associated with intellectual giftedness cannot be used as diagnostic elements; and even if they were, it is doubtful whether teachers would be very reliable judges. A risk is that gifted samples based on teacher nominations may reflect stereotypes rather than objective intellectual giftedness (Deku 2013; Siegle and Powell 2004). In essence, this method suffers from the same limitation of being a result of academic achievement, although in a subtler way than with the direct use of school grades. The use of psychometric tests to assess intellectual abilities is one of the most widely used criteria for inclusion in gifted groups and, in theour opinion, the most relevant. However, the threshold for recognizing intellectual giftedness differs from one study to another and from one specialized gifted program to another. An IQ threshold of 130 and above is theoretically accepted (Carman 2013; to go further, McBee and Makel 2019) and is in line with current criteria used to identify intellectual disability (two standard deviations above or below the mean). Yet unfortunately, the criterion of IQ ≥ 130 is rarely applied within studies on intellectual giftedness and associated non-cognitive characteristics. A related criterion used in some studies (Alnabhan (2011) and Tirri and Pehkonen (2002) for moral development; Li et al. (2017) for emotional intelligence; Guez et al. (2018) for success in academic examination) is to refer to a test assessing fluid intelligence, i.e., abstract reasoning abilities, as a criterion for inclusion within gifted groups. This idea is motivated by the fact that fluid intelligence and general intelligence (g) are very strongly correlated (Kovacs and Conway 2016). This approach is valuable because it provides a measure of intellectual abilities that is less biased by culture and verbal abilities highly dependent on socio-cultural and economic levels (Neff 1938). Nonverbal intelligence tests provide particularly useful information in fields of study where there is high verbal input, such as the study of moral development and emotional intelligence. In addition, non-cognitive abilities are sometimes assessed using tasks that are closely related to verbal comprehension and expression skills. To ensure the validity of studies of giftedness using these types of measures, the verbal level of participants must be carefully controlled to improve confidence in the conclusions that are drawn. Observed differences may then be related to the variable being measured, such as high intellectual ability or high moral skills, rather than to differences in verbal ability. Finally, some of the guidelines for identifying gifted students suggest that performance and non-performance identification methods can be treated as interchangeable. However, these gifted identification methods with and without performance tend to identify different students. For this reason, methods of gifted identification using ability tests or not cannot replace each other and should be used simultaneously (Acar et al. 2016).

2.2. Sampling Methods and Sample Sizes

In a research field such as that of non-cognitive characteristics related to intellectual giftedness, including psychological and socio-emotional dimensions, it seems particularly relevant to be careful regarding sampling methods and their consequences: socio-emotional features of a child may be determinant in whether they are academically successful and whether they are identified as gifted or not, in turn creating the possibility of a severe sampling bias. Unfortunately, when it conducting this literature review, it became apparent that the vast majority of participants included in studies of gifted samples were from gifted academic programs or summer camps for the gifted. This severely limits the extent to which findings can be generalized to the general population of gifted children and adolescents. Only certain profiles of gifted youth may be selected to participate in such programs—presumably those who are academically successful and socially well-adjusted. It could also be the case that these particular settings influence gifted children’s and adolescents’ development by changing their experience of being gifted and their self- and peer representations. Ideally, in order to capture the heterogeneous profiles of intellectually gifted youth, samples should be proportionally composed of participants actually representative of the intellectually gifted child and adolescent population. In particular, this would require broad inclusion of intellectually gifted participants not identified as intellectually gifted, and inclusion of highly gifted participants. Strictly adhering to these expectations would be overly ambitious, making it very difficult to conduct research on intellectual giftedness. A more realistic perspective in the our opinion would be to favor feasible sampling methods that depend on the objective of the study (to report a general phenomenon common to all intellectually gifted youth, or a context-dependent phenomenon?). For studies of non-cognitive characteristics associated with intellectual giftedness, a good solution regarding expectations and feasibility would be to use a mixed sample design. For example, it would be optimal to replicate the same results with participants from specialized programs for gifted students (specialized classes for gifted students, summer university, etc.), participants from regular classes not yet identified as gifted (and tested through screening), and/or participants from associative and/or clinical settings. It is fair to note, however, that access to some sampling methods strongly depends on which country the study is conducted in. For example, in the United States or Israel, specialized programs for the education of the gifted are widespread, but this is uncommon in other areas, such as West Europe. Systematic testing of children and systematic inclusion of gifted youth in specialized programs obviously facilitates the constitution of larger samples of gifted students (for example, see Lee and Olszewski-Kubilius 2006; Scholwinski and Reynolds 1985; Shechtman and Silektor 2012; and Zeidner and Schleyer 1999). Indeed, a recurring concern in research on giftedness is the difficulty in constituting large samples of gifted participants. ReIn thisearchers literature review, we find a median gifted sample size of 97 participants (min = 15; max = 1062). Based on studies included in this paper, it would be reasonable to recommend sample sizes of at least 100 participants for future studies as a way to encourage greater robustness and representativeness of the results. Another critical challenge for future studies in the field of non-cognitive characteristics in giftedness that stems from sampling considerations would be to fully consider the implications of cross-cultural comparisons (Hanushek 2021). Perceptions of giftedness can be vastly different from one place to another. In France for example, giftedness tends to carry a strong stigma and gifted children are often treated as nerds, i.e., as outsiders from the main group. The reverse can be true in other countries, potentially leading to very different socio-emotional consequences (e.g., a disadvantage versus an advantage in terms of socialization), and explaining inconsistent results in the literature. Some studies examined the impact of cultural differences (including participants’ family, educational, and social backgrounds) on the overall experience of giftedness (Levy and Plucker 2003; Rodgers 2008; Thomas 2008; and Yoon and Gentry 2009), and some studies focused on giftedness representations in various countries (Chan 2002; Coleman and Cross 2014; Lee-Hammond 1999; Neumeister et al. 2007; and Tavani et al. 2009). It would be helpful to complement these findings with cross-cultural studies focusing on the different elements addressed in this literature review (see e.g., Lee et al. 2020), which would provide more robust conclusions than the comparison of results from very different cultures as rwesearchers conducted here.

2.3. Effect Sizes

Effect sizes are needed to correctly interpret statistical results (Pek and Flora 2018), and are a requirement of APA norms (7th edition). This indicator is very often missing within this field of literature. In the present study, 39% of included studies did not provide effect sizes. It is important to recall here that an effect can be statistically significant while being of very small magnitude. This is critical to correctly interpreting differences between gifted and non-gifted youth: a significant difference between gifted children and their peers does not have the same implications for clinical practice depending on whether their scores are on average 1% or 50% higher than their peers. Observing large effects should not be an absolute prerequisite for publication (small effects can also hint at meaningful specificities of giftedness), but the level of transparency involved in reporting effect sizes in the results is strictly necessary and useful for future research. When effect sizes are available in this literature, they tend to be systematically very small. For the results synthesized in this literature review, the median percentage of variance explained by giftedness was 4.23% (min = 0%; max = 65.61%). As it stands, such results suggest negligible or no effect of giftedness in most of the domains investigated in this paper. This contributes to the conclusion that none of the domains considered here can be used as a diagnostic criterion for giftedness: apart from the fact that results tend to be unstable, giftedness generally does not explain enough variation in the measures to create a meaningful separation between gifted and non-gifted groups. It is also worth recalling that statistical power in a study depends on sample size and effect size. Considering the pre-existing findings on the topic and the generally very low effect sizes, future studies on non-cognitive characteristics of giftedness should favor large sample sizes to put forth robust findings.

2.4. Measurement Tools

A weakness in research on non-cognitive characteristics of the intellectually gifted is the overall lack of consistency when comparing results from different studies. Part of the problem may stem from the use of various measures and instruments that do not assess the same facets/components of a given concept. This statement applies to meta-analytic studies as well as to integrative reviews of works on anxiety, humor, moral development, and emotional intelligence. Another issue is that measurement tools sometimes display poor psychometric qualities. The extent to which results can be generalized to a larger group of individuals depends on the reliability and validity of the measurement (Scholwinski and Reynolds 1985). Moreover, the psychometric qualities of a measure depend on its context of use: reliability and validity are the properties of a particular scale in a particular population, not a property of the scale itself. In principle, the psychometric qualities of tests initially designed to assess non-gifted children and adolescents should therefore be examined (either before or after performing the study of interest) in a sample of intellectually gifted participants to verify that the structure and the qualities of the instrument are the same for the gifted and non-gifted groups. In practice, even researchers who do not want to engage in a full psychometric analysis (e.g., examination of measurement invariance across populations) can at least check the factor structure and the internal consistency of their instrument (e.g., Cronbach’s alpha) systematically and with little effort. It is also worth mentioning the difference between self-report and other-report measures. Some studies rely exclusively on parents and teacher-rated assessments. These methods can provide valuable information, but they are not suitable for all situations. Halo effects and confirmation bias are likely in this field of study, with many teachers and parents having firmly entrenched preconceptions of what gifted children are like (Lee-Hammond 1999; Neumeister et al. 2007; and Tavani et al. 2009). One way to overcome this difficulty is to use mixed-design studies whenever possible, including both self-report and other-report data (e.g., see Schuler 2000).