Seven Questions about Creativity and Creative Thinking

7. Does socio-economic disadvantage hinder students’ creativity?

The PISA 2022 results showed that, while everyone has the potential to be creative, differences in student socio-economic status may prevent many students from fully reaching their creative potential. On average, advantaged students scored close to one standard deviation higher than disadvantaged students in the PISA 2022 creative thinking test – enough to significantly decrease the overall mean performance of some countries and economies (OECD, 2024[1]). While student reading and mathematics scores explained a large proportion of the variation in students’ creative thinking scores, there were nonetheless significant differences in the performance of advantaged and disadvantaged students in creative thinking even after accounting for student performance in the core domains (OECD, 2024[1]). In general, research into the association between socio-economic status and creativity has tended to find a small but significant positive relationship (Acar et al., 2023[2]; Castillo-Vergara et al., 2018[3]).

There are many reasons why socio-economic disadvantage may hinder students’ creative potential. Beyond the well-documented economic, cultural and educational experiences and mechanisms that are known to affect student achievement overall, advantaged students may also have better access to diversified life experiences and activities that can nurture creative idea generation (Xu and Pang, 2020[4]; Jankowska, Lebuda and Gralewski, 2024[5]). Access to more diverse life experiences might support creative thinking in three ways: (1) by enabling students to develop a more heterogenous knowledge base from which to come up with creative ideas (Gocłowska and Crisp, 2014[6]); (2) by exposing students to new norms more frequently, in turn increasing cognitive flexibility (Ritter et al., 2012[7]); and (3) by increasing individuals’ openness to experience and intellect, personality traits that have consistently demonstrated a strong association with creative performance (Batey and Furnham, 2006[8]; Zhang et al., 2018[9]).

On the other hand, socio-economic hardships might expose students more frequently to the types of everyday problem-solving situations that require creative thinking (Acar et al., 2023[2]). Economic disadvantage and/or membership of a minority culture could be considered developmental adversities that provide an alternative type of diversifying experience, in line with the “hidden talents” theory (Ellis et al., 2023[10]), and research has found that children facing difficult socio-economic and cultural circumstances can develop enhanced cognitive abilities as an adaptive response (Young et al., 2022[11]).

Overall, while economic privilege can facilitate access to certain diverse life experiences, not all are inherently costly: for example, bilingualism and multiculturalism has been found to be positively associated with creativity (Lee and Kim, 2011[12]). Other factors outside of access to new experiences might also contribute to the poorer performance of disadvantaged students in creative tasks, including lower levels of self-efficacy (Lu, Ding and Nie, 2024[13]; Karwowski, 2011[14]) or challenges with executive functions like cognitive flexibility, working memory and inhibition that are known to support cognitive performance across tasks (Kupczyszyn, Filippetti and Oros, 2024[15]).

What new insights can data from the PISA CT Rescoring project shed on the relationship between creative thinking performance and socio-economic status? Do socio-economic differences persist across different scoring methods, and which aspects of coming up with creative ideas do disadvantaged students struggle with most? Do these patterns persist across country-language groups and different types of task? And what role (if any) do factors like multilingualism, self-efficacy or immigrant background play in mediating socio-economic differences in creative thinking performance?

7.1. Which scoring criteria do disadvantaged students struggle with most?

As expected, results from the PISA CT Rescoring project find that student socio-economic background is significantly associated with their scores derived from both the holistic judgement method and the criteria-based method (Box 7.1). On average across country-language groups, students’ standardised holistic creativity scores and mean response idea sum scores were positively correlated (r=0.2 and r=0.17, respectively) with their PISA economic, social and cultural (ESCS) index (Tables B7.1 and B7.2). Though the correlation with both scores is weak overall, it is nonetheless significant. Moreover, disadvantaged students – those in the bottom quartile of the PISA ESCS index within their country-language group sample – scored 0.75 points less than their advantaged peers (those in the within-group top quartile of the PISA ESCS index) on the standardised holistic creativity scale, on average, which represents half a standard deviation in performance (Table B7.3).

Box 7.1. Which score data are used in this chapter?

The analyses in this chapter draw upon two types of rescored data:

• Holistic creativity scores: Each student response was scored based on judges’ evaluations of the overall creative quality of the response. The holistic creativity score reflects a relative ranking of creative quality compared to all other responses in that country-language-task group. To account for differences in the distribution of scores across country-language-task groups, the analyses in this chapter standardised and transformed the raw holistic creativity scores (original range 1-7 score points) onto a scale that ranges from 1-10 score points, with a mean of 5 score points.

• Criteria scores: Each idea within a student response was scored for its appropriateness, originality and value. A cumulative idea sum score was also produced by aggregating the scores across these three criteria. Tasks classified as ‘generate diverse ideas’ tasks required students to submit multiple ideas per response; for these tasks, criteria scores across ideas were averaged so that each student response is associated with a single appropriateness, originality and value score.

Terms used in this chapter

• “Advantaged students”: Students that score in the top quartile of the PISA economic, social and cultural (ESCS) index, with respect to the country-language group sample included in the PISA CT Rescoring project.

• “Disadvantaged students”: Students that score in the bottom quartile of the PISA economic, social and cultural (ESCS) index, with respect to the country-language group sample included in the PISA CT Rescoring project.

This chapter focuses on examining differences in idea quality between advantaged and disadvantaged student groups. For more information on the scoring methods, see Annex A2.

The mean response appropriateness, originality and value scores of advantaged and disadvantaged students reveal interesting variation in the sources of this socio-economic gap in creative thinking performance. Figure 7.1 shows that the largest score gaps occur with respect to mean response appropriateness and flexibility scores: advantaged students score 0.3 score points and 0.2 score points higher than their disadvantaged peers on these criteria, respectively, on average across country-language groups (Table B7.4). The mean score gap for response originality and value scores are also just below 0.2 score points, though these results are significantly influenced by the removal of non-appropriate ideas from the analyses (see Box 7.2). These differences in task scores between advantaged and disadvantaged students represent around one-third of a standard deviation in performance for appropriateness scores compared to around one-fifth of a standard deviation in performance for originality and value scores.

Figure 7.1. Disadvantaged students struggle the most in coming up with appropriate ideas

Difference in mean response criteria scores between advantaged and disadvantaged students (all tasks), on average across country-language groups, before and after accounting for mathematics and reading

Note: All differences are significant (bars and markers). Bars represent mean score differences between advantaged and disadvantaged students before accounting for mathematics and reading scores, markers represent mean score differences after accounting for these factors. Mean score differences were computed on the pooled responses across all tasks in each country-language group, then averaged across country-language groups.

Source: Table B7.4

Box 7.2. Treatment of non-appropriate ideas for some analyses in this chapter

When calculating the mean response originality and value scores, only appropriate ideas were retained. According to the scoring methodology for the criteria-based approach (see Annex A2), judges were instructed to first score student ideas for appropriateness. Any ideas that achieved an appropriateness score of 0 were not scored further, meaning there are no originality or value scores associated with non-appropriate ideas. This process was followed to minimise the coding burden on judges and to simplify the interpretation of the criteria-based scores, following the argument that inappropriate ideas may well be original but they cannot be creative (Runco and Jaeger, 2012[16]).

Differences in mean response criteria scores between disadvantaged and advantaged students should be interpreted with this methodology in mind. For example, the mean appropriateness scores of advantaged and disadvantaged students are calculated using all idea scores, including those from students that scored 0 on idea appropriateness due to a lack of meaningful engagement with the task. In contrast, the mean score gap observed between advantaged and disadvantaged students for response originality and value, respectively, is calculated on only partially or fully appropriate ideas, and therefore likely excludes the most disengaged students from the sample in this analysis. As a result, it would be reasonable to expect the score difference between advantaged and disadvantaged students in mean response originality and value to be smaller than the equivalent in mean appropriateness. Indeed, when the originality and value scores of non-appropriate ideas are included in the analyses (recoded as 0 scores), then the score differences between advantaged and disadvantaged students are greater in mean response originality and value than in mean response appropriateness (Table B7.5).

The relatively large difference between advantaged and disadvantaged students in mean response appropriateness scores may be influenced by two factors. First, students from disadvantaged backgrounds may struggle more with the convergent thinking processes involved in creative thinking that support the evaluation and selection of ideas that are best suited for a given purpose. While the results from the PISA 2022 assessment found that the largest performance differences between advantaged and disadvantaged students were observed in the ‘generate diverse ideas’ tasks (OECD, 2024[1]), the PISA CT Rescoring study data suggest that disadvantaged students scored relatively worse in the ‘generate creative ideas’ tasks included in the study than they did in the ‘generate diverse ideas’ tasks (Figure 7.2). Our results in Chapter 5 of this report also broadly suggest that advantaged students may be more successful at generating creative ideas within more appropriate constraints than their disadvantaged peers. However, more research is needed to better understand potential differences in the cognitive processes involved in creative thinking between advantaged and disadvantaged students.

Figure 7.2. Disadvantaged students scored comparatively lower in ‘generate creative ideas’ tasks than in ‘generate diverse ideas’ tasks, relative to advantaged students

Mean response idea sum scores of advantaged and disadvantaged students, on average across country-language groups, by ideation process

Note: Average scores are computed by task first in each country-language group and then averaged across relevant tasks in each task group. The idea sum score refers to the sum of the idea appropriateness, originality and value scores. For 'generate diverse ideas' items that contain multiple ideas per response, the idea sum scores for all ideas were averaged to give a single mean idea sum score per student response.

Source: Table B2.2 and Table B7.6

The second factor that may play an even bigger influence in socio-economic differences in creative thinking performance is domain readiness. It stands to reason that proficiency in the core academic subjects provides students with a solid foundational knowledge and skills that are beneficial to creative thinking in a range of contexts, possibly explaining why socio-economic differences in creative thinking have been found to weaken with age and as students become more educated (Acar et al., 2023[2]). Indeed, students’ mathematics and reading scores accounted for a large proportion of the observed differences in creative thinking performance by socio-economic status in the PISA 2022 results (OECD, 2024[1]).

Do reading and mathematics scores similarly account for most of the observed performance gap between advantaged and disadvantaged students in the PISA CT Rescoring project results? Differences in standardised holistic creativity scores between advantaged and disadvantaged students reduce from 0.75 score points to just under 0.3 score points (i.e. from half a standard deviation in performance to around one-fifth), after accounting for student mathematics and reading scores (Table B7.3). While this difference remains significant, on average across country-language groups, it notably becomes insignificant in Canada (English and French provinces), Chile, Germany, France and Saudi Arabia. Figure 7.1 also shows that the performance gap on each of the criteria-based scores is significantly reduced after accounting for student mathematics and reading scores: differences between advantaged and disadvantaged students in each of the score criteria fall to below 0.1 score points, on average across country-language groups. While these differences are small and often insignificant at the individual country-language group level, they remain significant on average (Table B7.4).

Overall, the much smaller gaps in creative thinking scores after accounting for students’ reading and mathematics scores is cause for optimism, highlighting that a significant proportion of the socio-economic differences in creative thinking appear remediable with more equitable teaching and learning opportunities for students. However, significant differences still exist even after accounting, meaning that other factors beyond foundational literacies contribute to socio-economic disparities in creative thinking.

7.1.1. Are there cross-cultural differences in the relationship between socio-economic status and creative thinking scores?

The association between socio-economic status and creative thinking scores varies across country-language groups. The standardised holistic creativity scores showed a stronger association with student socio-economic status across all country-language groups (Figure 7.3), perhaps because this scoring method captures or is influenced by broader aspects of student responses that might also be associated with socio-economic status (e.g. vocabulary, spelling and grammar, etc.). Additionally, the holistic judgement method evaluates student responses relative to their peers, and this forced comparison may contribute to the stronger association observed between holistic creativity scores and socio-economic status (i.e. regardless of large or small differences in overall idea quality, disadvantaged students tend to provide comparatively less creative answers than their advantaged peers).

The strongest associations between students’ standardised holistic creativity scores and their socio-economic status were observed in Brazil, Colombia, Portugal and Greece, with over 5% of the variation in scores in Portugal and Greece and over 7% of the variation in scores in Colombia and Brazil explained by students’ socio-economic index (Figure 7.3). In contrast, less than 3% of the variation in holistic creativity scores can be explained by socio-economic status in Hong Kong (China), Canada (English and French provinces) and Saudi Arabia. The criteria-based method follows a largely similar pattern, albeit the proportion of variation in students’ mean response idea sum scores explained by socio-economic status is smaller in all country-language groups except for Canada (French provinces).

Despite differences across country-language groups, students with a higher socio-economic status score significantly higher than students with a lower socio-economic status, on average: a one-unit increase in the PISA ESCS index is associated with a difference of 0.36 standardised holistic creativity score points, which represents around a quarter of the standard deviation in performance amongst students. In Brazil and Colombia, the score difference is even larger – closer to a third of the standard deviation in performance. It is unsurprising that countries with greater within-country inequality in student socio-economic and cultural status provide relatively more educational opportunities to privileged students, which in turn may act to widen performance disparities in creative thinking. Indeed, Colombia, Brazil and Portugal all have significantly larger within-country variation in student socio-economic status than other country-language groups that participated in the PISA Rescoring project (OECD, 2023[17]).

Figure 7.3. Socio-economic status explains a larger proportion of the variation in holistic creativity scores than students’ mean response idea sum scores

Proportion of variation in holistic creativity scores (bar) and mean response idea sum scores (marker) explained by socio-economic status (all tasks), by country-language group

Note: Significant r-squared values (i.e. proportion of explained variation) are shown by bars/markers in a darker shade. Country-language groups are ordered in descending order of the proportion of explained variation in holistic creativity scores.

Source: Table B7.8 and Table B7.9

As discussed in the previous section, differences in the relationship between socio-economic status and creative thinking scores in each country-language group is also likely influenced by the equivalent relationship between academic achievement and socio-economic status more broadly (Dai et al., 2012[18]). Some countries and economies with a stronger association between socio-economic status and mathematics performance (OECD, 2023[17]), for example Belgium, France and Colombia, also show above-average variance in creative thinking scores explained by socio-economic status. After accounting for both mathematics and reading scores, the influence of socio-economic status on students’ standardised holistic creativity scores reduces substantially but remains significant in all country-language groups except Chile, Germany, France and Saudi Arabia (Table B7.10).

7.2. Are socioeconomic differences more pronounced in some types of task?

The results so far have shown that the association between socio-economic status and creative thinking performance persists across country-language groups and scoring methods, and that disadvantaged students struggled relatively more in the ‘generate creative ideas’ tasks than in the ‘generate diverse ideas’ tasks included in the PISA CT Rescoring study (Figure 7.2). Looking at the scores of students on a task-by-task basis, Figure 7.4 shows that the variation in creativity scores explained by socio-economic status ranges across tasks – from around 3% of the variation in Task 2 (the Illustration Titles task; see Figure C in the Introduction) to close to 8% of the variation in Task 7 (The Exhibit; see Figure H).

Figure 7.4. Tasks differ in how much socio-economic status influences student creativity scores

Proportion of variation in holistic creativity scores (bar) and mean response idea sum scores (marker) explained by socio-economic status, by task (on average across country-language groups)

Note: All r-squared values (i.e. proportion of explained variation) are significant. Tasks are ordered in descending order of the proportion of explained variance in holistic creativity scores.

Source: Table B7.8 and Table B7.9.

There are mixed findings in previous research on the effects of task modality and socio-economic differences in performance in creative thinking tasks. Results from the PISA 2022 assessment, as well as other empirical studies, suggest that verbal tasks (i.e. tasks requiring writing) tend to exacerbate socio-economic differences in performance compared to those that require students to demonstrate creativity non-verbally (OECD, 2024[1]; Gajda, Karwowski and Beghetto, 2017[19]). However, a recent meta-analysis found that there were no significant differences in the performance of students with different socio-economic status on verbal and non-verbal creativity tasks (Acar et al., 2023[2]).

In the PISA CT Rescoring study results, when looking at the average character count of student responses to different tasks, there is an overlap between the average verbosity of tasks and the strength of the association between student scores and socio-economic status on the task (see Table 7.1). For example, the three tasks with the strongest association between student scores and socio-economic status place significant writing demands on students compared to the other tasks included in the study. In The Exhibit (Task 7) students are asked to describe three elaborate engineering solutions for a problem; the task 2983 (Task 3; see Figure D in the Introduction) asks students to describe an idea for a story; and the Food Waste task (Task 5; see Figure F) requires students to describe three solutions to address supermarket waste. Conversely, both the Science Fair Poster and Illustration Titles tasks (Tasks 1 and 2; see Figures B and C) require limited writing from students and show much weaker associations with socio-economic status and student scores on average across country-language groups.

Table 7.1. Some tasks have higher response demands than others

Mean response character count by task, on average across country-language groups

Task number	Task title	Mean response character count
Task 2	Illustration Titles	30
Task 1	Science Fair Poster	71
Task 4	Save the Bees	78
Task 5	Food Waste	85
Task 6	Save the River	101
Task 7	The Exhibit	110
Task 3	2983	203

Note: Tasks are ordered in ascending order of average character count (lowest to highest). In the Science Fair Poster task, the character count refers to the optional poster description response box that students can choose to complete to accompany their poster design. Mean response character counts are rounded to the nearest integer.

Source: Table B7.12

While mean character count alone does not necessarily equate to task demand, nor to the verbal proficiency required to respond sufficiently to a given task, being able to successfully convey one’s ideas will inevitably affect the extent to which judges understand and value them – especially in tasks that require more elaborate responses. Although judges were explicitly instructed to focus on response content rather than grammatical form or style, it may be that judges were influenced by these factors or the level of detail provided within student responses. After accounting for students’ mathematics and reading scores, the association between students’ creativity scores and their socio-economic status considerably reduces in each task, though it remains significant across all tasks (Table B7.10).

Task context and instructions might also have a significant effect on how creativity tasks are approached by different student groups. For example, the limits of an appropriate response are likely more constrained within an engineering problem-solving task compared to a graphic design task or creative writing exercise, which may in turn create the impression or expectation that “correct” answers exist in some types of task. Indeed, the results in the previous section of this chapter show that appropriateness was the most challenging criterion for disadvantaged students to achieve, and the results in Chapter 1 show that response appropriateness tends to be more correlated with creative scores in the problem-solving tasks than the expressive tasks. Less confident students may therefore struggle to think of creative ideas in more practical problem-solving tasks for fear of not providing a “good enough” answer (see next section).

7.3. What factors might moderate the relationship between socio-economic status and creative thinking performance?

7.3.1. Self-efficacy beliefs

Some research has suggested that socio-economic differences in creative thinking performance may be mediated by the testing environment itself (Acar et al., 2023[2]). When a task is understood as a test, it can activate self-directed beliefs that shape performance (Spencer and Castano, 2007[20]). Students from disadvantaged socio-economic backgrounds are more likely to hold unfavourable beliefs towards their own creative abilities, leading to lower creative scores through less engagement and risk-taking (Yang et al., 2020[21]; Beghetto, Karwowski and Reiter-Palmon, 2021[22]; Puente-Díaz, 2016[23]).

Our data shows a positive but weak correlation between students’ standardised holistic creativity scores and their creative self-efficacy and mathematics self-efficacy, respectively, on average across country-language groups (Table B7.13). Figure 7.5 shows that the gap in standardised holistic creativity scores between advantaged and disadvantaged students reduces after accounting for students’ self-efficacy beliefs across all country-language groups, albeit to varying degrees. The largest reductions in score differences were observed in Belgium (French community) (0.35 score points), Greece and Germany (both 0.24 score points), in contrast to only small differences in the score gap observed in Hong Kong (China) and Colombia. Despite moderating the relationship between socio-economic status and creative thinking scores to some degree in each country-language group, belonging to the highest socio-economic quartile and not the lowest remains a significant predictor of success in creative thinking across the board, even after accounting for students’ self-efficacy beliefs.

Figure 7.5. Self-efficacy beliefs moderate socio-economic disparities in standardised holistic creativity scores in all country-language groups

Mean difference in standardised holistic creativity scores between advantaged and disadvantaged students (all tasks) before and after accounting for self-efficacy beliefs, by country-language group

Note: All differences (before and after accounting for self-efficacy beliefs) are significant. Country-language groups are ordered by the largest reduction in the difference in standardised holistic creativity scores between advantaged and disadvantaged students after accounting for student self-efficacy beliefs (creative self-efficacy and mathematics self-efficacy). Mean score differences between advantaged and disadvantaged students are computed on the pooled responses across tasks in each country-language group.

Source: Table B7.14

The role of self-efficacy beliefs in moderating socio-economic differences in performance also varies by task. The score difference (holistic creativity scores) between advantaged and students in The Exhibit and the Save the River tasks – the two scientific problem-solving tasks – reduces by close to 0.2 score points, after accounting for students’ creative and mathematics self-efficacy beliefs (Table B7.15). Across the other tasks included in the study the relative score difference after accounting for self-efficacy beliefs also reduces, though to a lesser degree (between 0.12 and 0.16 score points).

These findings on self-efficacy, creative thinking scores and student background suggest that cultivating self-efficacy beliefs is important for supporting creative performance in all types of tasks. Helping students to feel confident in both their creative and general academic abilities may be especially important for supporting students from disadvantaged backgrounds in tasks that are more difficult and/or that have a more explicit focus on domain knowledge and skills (like scientific creativity).

7.3.2. Diverse life experiences: Multilingualism and immigration status

As described in the introduction to this chapter, diverse life experiences may support creative thinking skills. While socio-economic advantage can facilitate access to a range of experiences and opportunities, not all diverse life experiences are inherently costly. Being fluent in multiple languages and immersion in different cultures, for example, have been independently shown to support creative potential (Fürst and Grin, 2021[24]).These experiences can enhance executive functions and cognitive flexibility, enabling individuals to recognise different perspectives and create meaningful associations between diverse concepts, which in turn strengthen fluency, elaboration, originality and flexibility in creative tasks (Ghonsooly and Showqi, 2012[25]; Tadmor, Tetlock and Peng, 2009[26]; Tadmor, Galinsky and Maddux, 2012[27]). While language fluency and exposure to different cultures might be developed through travel, it can also be facilitated through living in different countries and/or having family members from a different socio-cultural or linguistic background.

Findings from the PISA CT Rescoring study show that student immigrant status can have a mixed influence on creative thinking scores of students. Figure 7.6 shows students’ mean standardised holistic creativity scores by immigrant background group (no immigrant background, first-generation, second-generation). On average across country-language groups, students with no immigrant background have a small but statistically significant advantage over their peers with an immigrant background (score difference of around 0.2 points, or less than a sixth of a standard deviation in scores). However, the size and direction of score differences between immigrant and non-immigrant background students vary across country-language groups: score differences in favour of non-immigrant background students were largest in Slovenia, Chile, Belgium (French community), France, Germany and Greece; statistically insignificant in Italy, Hong Kong (China), Portugal, Canada (English and French provinces); and reversed in Saudi Arabia (students with an immigrant background scored nearly 0.5 score points more than those with no immigrant background, or around one third of a standard deviation higher).

Figure 7.6. Students with an immigrant background tend to achieve lower scores than students without an immigrant background

Mean standardised holistic creativity scores by immigrant background status (all tasks), by country-language group

Note: Statistically significant differences between immigrant (all) and non-immigrant students are shown by filled markers. Country-language groups are ordered by the mean difference in standardised holistic creativity scores between immigrant (all) and non-immigrant students (non-immigrant – immigrant) (highest to lowest). Country-language groups with fewer than 20 first- or second-generation immigrant students in the sample were excluded from the analysis. Mean score differences are computed on the pooled responses across tasks in each country-language group.

Source: Table B7.16

Differences in creativity scores between students with an immigrant background and those without an immigrant background seem to be largely influenced by differences in response appropriateness scores, which in turn appear driven by mathematics and reading scores. Figure 7.7 shows that in six of the 12 country-language groups included in the analysis, the largest difference in response quality between immigrant and non-immigrant-background students relates to response appropriateness scores, with students with an immigrant background scoring around 0.1 score points lower on average across country-language groups than students with no immigrant background.

Figure 7.7. Mathematics and reading scores explain most of the differences in scores by immigrant background

Differences in mean response criteria score (appropriate, original and value) between non-immigrant and immigrant background students (all tasks) before and after accounting for reading and mathematics scores

Note: Significant differences are shown by colours in a darker shade for each criteria (bars) before accounting, and by markers of a darker shade after accounting Mean score differences were computed on the pooled responses across all tasks in each country-language group.

Source: Table B7.17

While these results might be driven in part by the treatment of non-appropriate responses when comparing originality and value scores between student groups (see Box 7.2), they also appear closely related to differences in students’ reading and mathematics proficiency. After accounting for student scores in the core domains, immigrant students in many country-language groups tended to perform no differently to their peers without an immigrant background across the different score criteria. On average, and in Belgium (French community), Canada (French provinces), France and Slovenia, the response appropriateness scores of students with an immigrant background even reversed from being significantly lower than their non-immigrant background peers to significantly higher than their peers, after accounting for reading and mathematics scores. The same reversal after accounting for mathematics and reading was also observed in Greece for response value scores and Slovenia for response originality scores, with immigrant background students significantly outperforming their non-immigrant counterparts, and in Saudi Arabia, where non-immigrant background students then significantly outperform their immigrant background peers in response originality and value scores.

Overall, after accounting for students’ reading and mathematics scores, students with an immigrant background tend to score slightly but significantly higher than students without an immigrant background – both in their standardised holistic creativity scores and in their mean response idea sum scores. Taken all together, these findings suggest that students with an immigrant background tend to score lower on the creative thinking tasks not because of their immigrant background per se, but perhaps rather due to relative difficulties in understanding and expressing themselves in the test language. When students with similar reading proficiency are compared, immigrant students perform no worse than non-immigrant students in response originality and value scores, and even significantly outperform their non-immigrant peers in response appropriateness and in overall response quality (standardised holistic creativity scores and mean response idea sum scores), on average.

Language proficiency in multiple languages is also associated with higher creativity scores across the two scoring methods used in this chapter (Table B7.17). Students that can speak two or more languages well enough to converse with others scored significantly higher than students that speak only one language (a difference of 0.2 score points on the standardised holistic creativity scale and mean response idea sum scores, respectively), on average across country-language groups. However, after accounting for mathematics and reading scores, there are no significant differences in mean response idea sum score between students that can speak more than one language and those that cannot, and students that speak only one language in fact achieve higher standardised creativity scores (difference of 0.1 score points) after accounting for these factors. It may be that students who are more likely to be able to converse with others in multiple languages are those students that perform better academically, in general, and this largely accounts for the significant differences in performance observed before accounting for reading and mathematics performance.

In sum, the findings are mixed when it comes to the relationship between diverse life experiences and their relationship to creative thinking. While the results suggest that factors like immigration background and multi-linguicism do not adversely affect creative thinking, they also find that most of the difference in performance between student groups (positive or negative) is better explained by differences in reading and mathematics scores. Of course, it may also be that immigrant background and language proficiency, as measured in PISA (i.e. student questionnaire self-reports), are not the best proxies of diverse life experience, and/or that other diverse life experiences and opportunities might better mediate differences in creative thinking performance.

Key findings and implications

• Student socio-economic status and scores on the creative thinking tasks are significantly positively associated, in general, though the strength of the relationship is weak (r=0.2). However, a one-unit increase in the PISA ESCS index is associated with a significant increase in standardised holistic creativity score of 0.36 points (around a quarter of a standard deviation in performance).

  • The association between socio-economic status and creative thinking performance was strongest in countries with the largest within-country inequality (e.g. Colombia, Brazil and Portugal). It may be that privileged students in these countries have relatively better access to quality educational opportunities than their disadvantaged peers, which in turn widens performance disparities.

• Disadvantaged students scored 0.75 points less than advantaged students (around half a standard deviation in holistic creativity scores), on average. Disadvantaged students also scored lower in each of the criteria-based scores than advantaged students, particularly response appropriateness and flexibility.

• Differences in reading and mathematics proficiency between advantaged and disadvantaged student groups largely account for their performance differences in creative thinking.

  • Policy efforts should focus on providing more equitable teaching and learning experiences, and opportunities to practice creative thinking, for all students.

  • However, not all of the difference in creative thinking scores can be explained by academic performance – meaning that other factors beyond foundational literacies also contribute to socio-economic differences in performance.

• Disadvantaged students struggled more in certain types of creative thinking tasks, in particular convergent thinking tasks, tasks with greater writing demands, and problem-solving tasks:

  • Students from disadvantaged backgrounds may struggle with the convergent thinking processes involved in creative thinking that support the evaluation and selection of ideas that are best suited for a given purpose;

  • Disadvantaged students may also struggle to successfully convey their ideas, in part given their lower proficiency in core literacies (on average), or due to greater disengagement with tasks with higher response demands;

  • Given differences in the scores of students in the PISA core domains of reading and mathematics, disadvantaged students may have less “domain readiness” than advantaged students, i.e. relevant knowledge and skills related to a given problem context. Moreover, problem-solving tasks tend to be more constrained in terms of appropriateness than expressive tasks, meaning it is relatively harder to come up with original and valuable ideas that are also task appropriate.

• Self-efficacy beliefs effectively moderate students’ success in creative thinking tasks in all country-language groups, and especially in scientific problem-solving tasks (on average across country-language groups).

References

[2] Acar, S. et al. (2023), “Socio-economic status and creativity: A meta-analysis”, The Journal of Creative Behavior, Vol. 57/1, pp. 138-172, https://doi.org/10.1002/jocb.568.

[8] Batey, M. and A. Furnham (2006), “Creativity, intelligence, and personality: A critical review of the scattered literature”, Genetic, Social and General Psychology Monographs, Vol. 132/4, pp. 355-429.

[22] Beghetto, R., M. Karwowski and R. Reiter-Palmon (2021), “Intellectual risk taking: A moderating link between creative confidence and creative behavior?”, Psychology of Aesthetics, Creativity, and the Arts, Vol. 15/4, pp. 637-644, https://doi.org/10.1037/aca0000323.

[3] Castillo-Vergara, M. et al. (2018), “Does socioeconomic status influence student creativity?”, Thinking Skills and Creativity, Vol. 29, pp. 142-152, https://doi.org/10.1016/j.tsc.2018.07.005.

[18] Dai, D. et al. (2012), “Influences of social and educational environments on creativity during adolescence: Does SES matter?”, Creativity Research Journal, Vol. 24/2-3, pp. 191-199, https://doi.org/10.1080/10400419.2012.677338.

[10] Ellis, B. et al. (2023), The hidden talents framework: Implications for science, policy and practice, Cambridge Univeristy Press, https://doi.org/10.1017/9781009350051.

[24] Fürst, G. and F. Grin (2021), “Multicultural experience and multilingualism as predictors of creativity”, International Journal of Bilingualism, Vol. 25/5, pp. 1486-1494, https://doi.org/10.1177/13670069211019468.

[19] Gajda, A., M. Karwowski and R. Beghetto (2017), “Creativity and academic achievement: A meta-analysis”, Journal of Educational Psychology, Vol. 109/2, pp. 269-299, https://doi.org/10.1037/edu0000133.

[25] Ghonsooly, B. and S. Showqi (2012), “The effects of foreign language learning on creativity”, English Language Teaching, Vol. 5/4, pp. 161-167, https://doi.org/10.5539/elt.v5n4p161.

[6] Gocłowska, M. and R. Crisp (2014), “How dual-identity processes foster creativity”, Review of General Psychology, Vol. 18/3, pp. 216-236, https://doi.org/10.1037/gpr0000008216.

[5] Jankowska, D., I. Lebuda and J. Gralewski (2024), “Creating home: Socioeconomic status and home environment as predictors of family climate for creativity”, Thinking Skills and Creativity, Vol. 52, pp. 1-13, https://doi.org/10.1016/j.tsc.2024.101511.

[14] Karwowski, M. (2011), “It doesn’t hurt to ask... but sometimes it hurts to believe: Polish students’ creative self-efficacy and its predictors”, Psychology of Aesthetics, Creativity, and the Arts, Vol. 5/2, pp. 154-164, https://doi.org/10.1037/A0021427.

[15] Kupczyszyn, K., V. Filippetti and L. Oros (2024), “Socioeconomic status effects on children’s creativity: The mediating role of executive functions”, Thinking Skills and Creativity, Vol. 51, pp. 1-15, https://doi.org/10.1016/j.tsc.2023.101437.

[12] Lee, H. and K. Kim (2011), “Can speaking more languages enhance your creativity? Relationship between bilingualism and creative potential among Korean American students with multicultural link”, Personality and Individual Differences, Vol. 50/8, pp. 1186-1190, https://doi.org/10.1016/j.paid.2011.01.039.

[13] Lu, Z., Y. Ding and Y. Nie (2024), “How does family socioeconomic status affect creativity? The role of creative self-efficacy and critical thinking disposition”, Current Psychology, Vol. 43/6, pp. 5674-5681, https://doi.org/10.1007/s12144-023-04768-x.

[1] OECD (2024), PISA 2022 Results (Volume III): Creative Minds, Creative Schools, PISA, OECD Publishing, https://doi.org/10.1787/765ee8c2-en.

[17] OECD (2023), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, OECD Publishing, Paris, https://doi.org/10.1787/53f23881-en.

[23] Puente-Díaz, R. (2016), “Creative self-efficacy: An exploration of its antecedents, consequences, and applied Implications”, The Journal of Psychology, Vol. 150/2, pp. 175-195, https://doi.org/10.1080/00223980.2015.1051498.

[7] Ritter, S. et al. (2012), “Diversifying experiences enhance cognitive flexibility”, Journal of Experimental Social Psychology, Vol. 48/4, pp. 961-964, https://doi.org/10.1016/j.jesp.2012.02.009.

[16] Runco, M. and J. Jaeger (2012), “The standard definition of creativity”, Creativity Research Journal, Vol. 24/1, pp. 92-96, https://doi.org/10.1080/10400419.2012.650092.

[20] Spencer, B. and E. Castano (2007), “Social class is dead. Long live social class! Stereotype threat among low socioeconomic status individuals”, Social Justice Research, Vol. 20/4, pp. 418-432, https://doi.org/10.1007/s11211-007-0047-7.

[27] Tadmor, C., A. Galinsky and W. Maddux (2012), “Getting the most out of living abroad: Biculturalism and integrative complexity as key drivers of creative and professional success”, Journal of Personality and Social Psychology, Vol. 103/3, pp. 520-542, https://doi.org/10.1037/A0029360.

[26] Tadmor, C., P. Tetlock and K. Peng (2009), “Acculturation strategies and integrative complexity: The cognitive implications of biculturalism”, Journal of Cross-Cultural Psychology, Vol. 40/1, pp. 105-139, https://doi.org/10.1177/0022022108326279.

[4] Xu, X. and W. Pang (2020), “Reading thousands of books and traveling thousands of miles: Diversity of life experience mediates the relationship between family SES and creativity”, Scandinavian Journal of Psychology, Vol. 61/2, pp. 177-182, https://doi.org/10.1111/sjop.12591.

[21] Yang, Y. et al. (2020), “Hope and creative self-efficacy as sequential mediators in the relationship between family socioeconomic status and creativity”, Frontiers in Psychology, Vol. 11/438, pp. 1-8, https://doi.org/10.3389/fpsyg.2020.00438.

[11] Young, E. et al. (2022), “Hidden talents in context: Cognitive performance with abstract versus ecological stimuli among adversity-exposed youth”, Child Development, Vol. 93/5, pp. 1493-1510, https://doi.org/10.1111/cdev.13766.

[9] Zhang, D. et al. (2018), “Family socio-economic status and parent-child relationships are associated with the social creativity of elementary school children: The mediating role of personality traits”, Journal of Child and Family Studies, Vol. 27/9, pp. 2999-3007, https://doi.org/10.1007/s10826-018-1130-4.