Single-sex vs. Coeducational schooling: an empirical study on the mental health outcomes of middle school students

ABSTRACT

Unlike previous literature that mainly examined the cognitive effects of single-sex education, this study investigates its effects on students’ mental health, focusing on depression, suicide-related behaviours, and the various causes of stress. Utilizing a national-level large-scale dataset of middle school students in South Korea and employing an instrument variable approach, we find that attending single-sex schools reduces the probability of depression in female students by about 9.9% points, while no significant effect is observed for male students. Additionally, our research regarding the causes of stress uncovers that single-sex schooling reduces the stress associated with peer relationships, particularly for girls. These findings imply that the non-cognitive benefits of single-sex education may be stronger than previously thought. As a result, we advocate for a more thoughtful and inclusive implementation of school policies, taking into consideration the impact of gender peer effects at school.

KEYWORDS:

• Single-sex schooling
• mental health
• school formation
• gender peer effect

JEL CLASSIFICATION:

• I10
• I21
• I31
• J16

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The microdata of the KYRBS are available in the public domain: https://www.kdca.go.kr/yhs/.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/00036846.2023.2291096

Notes

¹ As of the 2014–2015 school year, 283 public schools qualified as single-sex schools after excluding juvenile justice facilities and alternative, special education, and vocational schools (Mitchell et al. 2017). In contrast, in March 2002, only around 12 public schools provided single-sex classrooms (Benham et al. 2019).

² In both Doe v. Vermilion Parish School Board in 2011 and Doe v. Wood County Board of Education in 2012, federal district courts ruled that the school districts had failed to meet the requirement of single-sex classes being entirely voluntary. Additional legal disputes include Reach Academy for Boys & Girls, Inc. v. Delaware Department of Education in 2014, where a federal district court granted parents’ requests for a preliminary injunction regarding a single-sex program (Benham et al. 2019; Eckes 2021).

³ Active associations include the Alliance of Girls’ Schools Australasia in the Australasia region, the National Coalition of Girls’ Schools in the North America region, and the Girls’ Schools Association in the United Kingdom.

⁴ Source: OECD, PISA 2018 Database, Table III.B1.3.11 (Programme orientation and gender composition at school).

⁵ The reports and microdata of the KYRBS are released annually in December each year (https://www.kdca.go.kr/yhs/). It’s worth noting that this study’s sample includes students from 400 middle schools on an annual basis. However, these schools change yearly based on shifts in student populations within school districts. The names of the participating schools are not made public due to official privacy concerns (Choi, Park, and Behrman 2015).

⁶ The yearbooks are obtained from the Korean Education Statistics Service website, https://kess.kedi.re.kr/eng/index.

⁷ Based on the sample means of the additional outcome variables summarized in Table A2, only 6% of students have thought about a suicidal plan, only 1% have visited hospital after attempting suicide, and only 3% have visited hospital due to violence victimization, while, on average, 61% of students responded that they feel happy.

⁸ The PSS and the PGSS have different variations. For example, the national average of the PGSS (PSS) is 47.67% (22.47%) in 2018. The highest PGSS (PSS) among 17 provinces is 50% (41.79%) and the lowest PGSS (PSS) is 42.11% (9.25%) in 2018.

⁹ This finding is consistent with the findings of previous studies (Gee and Cho 2014) that physical-violence victimization among same sex peers in schools is likely to be higher for male rather than female students.