An exploratory study of creativity, personality and schooling achievement

ABSTRACT

We investigate the link between schooling achievement and creativity scores, controlling for personality traits and other individual characteristics. Our study is based on field data collected in a secondary school situated in a Parisian suburb. Four scores of creativity were measured on 9th graders. Verbal divergent thinking negatively predicts the grades in most subjects, but graphical integrative thinking is positively correlated with scientific grades. There is no significant correlation with the other measures of creativity, implying a low importance of creativity in school. In line with previous work, we find that conscientiousness and openness are positively associated with grades.

KEYWORDS:

• Schooling achievement
• creativity
• personality traits
• field data

Acknowledgements

We are grateful to José de Valverde and the secondary public school ‘Saint-Exupéry’ of Rosny-Sous-Bois, France. We would also like to thank the anonymous referees from Education Economics, Colin Green, Louis Levy-Garboua, Todd Lubart, Jean-Christophe Vergnaud and Marie-Claire Villeval for their useful comments that contributed to improving the paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. In fact, we hypothesize here that there is a positive correlation between wages and school grades, through higher education attainment (Altonji 1995; Weiss 1995).

2. See Ai (1999), Balgiu and Adîr (2014), Cicirelli (1965, 1966), Furnham, Zhang, and Chamorro-Premuzic (2006), Jackson (2013), Nami, Marsooli, and Ashouri (2014), Sen and Hagtvet (1993), Yeh (2004).

3. Studies that try to isolate predictors of academic success generally use the Grade Point Average (GPA), which is an average score based on different school subjects (language, mathematics, literature, chemistry, etc.). ‘Secondary subjects’ such as music, arts or physical education are not included (Laidra, Pullmann, and Allik 2007; Richardson, Abraham, and Bond 2012; Steinmayr and Spinath 2008). The GPA cannot thus be a total representation of schooling achievement as it excludes some of the teaching. Despite criticism of the accuracy and validity of this index (Didier et al. 2006; Johnson 2003), it remains the most widely used assessment (Richardson, Abraham, and Bond 2012).

4. We favour test scores over IQ level as they represent the real decisional component of educational choices. In fact, the decision to invest in an additional year of schooling mainly depends on test scores.

5. This report can be found in French at http://ife.ens-lyon.fr/vst/DA-Veille/70-janvier-2012.pdf.

6. As opposed to cognitive factors, conative factors refer to personality traits and to motivation.

7. See Poropat (2009), Richardson, Abraham, and Bond (2012), Bowles and Gintis (1975), Bowles, Gintis, and Osborne (2001), Heckman (2006), Heckman, Stixrud, and Urzua (2006).

8. A relatively low proportion of our sample (24%) was involved in cultural activities such as playing an instrument. Field data could overcome this issue if we had the parents to fill in some questionnaires as well. Unfortunately, this is not the case. So we cannot exclude unobserved factors in our error terms.

9. We do not take into consideration here the stereotypes that girls and boys may encounter when choosing specialities.

10. In a dynamic setting, the following year's test scores () could be explained the grades at and the same cognitive and non-cognitive abilities.

11. A different version (A and B) was used in each session in order to obtain robust scores of creativity. The tasks were the same overall but the content differed (types of drawing, titles of the stories).

12. We allow pupils who switch three times to be consistent, considering that they might be indifferent between the first and third switch. For these individuals, their certainty equivalent is situated between the first and third row.

13. Each teacher (of the different subjects) comes up with their own tests. For our analysis, we standardized the test scores. For every school grade and subject, there is a specific programme to follow established by the French Ministry of National Education. It is known by all teachers and it is their duty to follow it so we can assume that tests are standard.

14. Hoxby (2000) and Markman et al. (2003) found that peer achievement has a positive effect on students' own achievement.

15. Even though we did not obtain some pupils' grades, they still took the exam. The mean average given is on a 40-point basis that can easily be converted to a 20-point scale. An average grade of 20/40, based on continuous assessment and exam grades, is required to pass this exam. Scoring 0/40 in one of the subject tests leads to immediate failure.

16. This information is available on http://www.education.gouv.fr/cid59753/diplome-national-du-brevet.html.

17. This is an indication of the school level being rather low.

18. There are three distinctions for the French diploma: the lowest (cum laude) is awarded for an average grade between 12/20 and 14/20, the second (magna cum laude) for an average grade between 14/20 and 16/20 and the highest (summa cum laude) for an average grade higher than 16/20.

19. However, a two-tailed t-test on the score of conscientiousness between those who attended the creativity sessions and those who did not yielded a p-value = .39.

20. Two-tailed t-tests on the standardized grades for each subject yielded a p-value < .1 except for biology and sports.

21. His study is based on 607 teenagers.

22. There are no significant trends over the year meaning that pupils overall exhibit constant grades during the school year. Thus, we can compute the year's average test scores.

23. We also ran the same regressions on the test scores for each subject but we found that it did not change the main results so, to have greater power in our data, we pooled them by field of study.

24. They use scales to evaluate various strategies such as cognitive learning strategies, or resource-related leaning strategies.

25. The former study was run on Israeli data and the latter was carried out in Norway.

26. We did not have access to their exact date of birth but the sessions during which we asked their age were run on January 24th and 25th. The distribution of age is the following: 2.7% of our sample is 13, 64.0% is 14, 28.0% is 15 and 5.3% is 16. The 16-year-olds and most of the 15-year-old pupils are likely to be repeaters.

27. These coefficients are not very stable because of our limited sample but still give an idea of what can occur in this situation.

28. These decisions might, however, be influenced by the exam results ex post, but this concerns a marginal number of pupils.

29. In fact, it was mandatory to pass the middle school exam, and it is still mandatory to pass the high school exam in order to reach higher educational levels.

30. Available upon request.

31. We asked pupils the socio-professional category of their parents as well as their exact job. We were then able to establish the proportion of highly qualified mothers and fathers in our sample, which is 35.4% and 29.1%, respectively.

Additional information

Funding

This work was supported by the French Research National Agency (ANR) under Grant number ANR-10-CREA-008.