Perry, Laura B.
School Composition and Student Outcomes: A Review of Emerging Areas of Research
Review of current research on school composition and its effects on student outcomes.
Journal Name or Institutional Affiliation:
Paper Presented at conference
Australian Association for Research in Education, November 2007 in Fremantle, WA
- PISA confirmed that SES affects student achievement, but the degree to which it does varies by country.
- Students perform better in schools that have a higher mean school SES, regardless of their individual SES. Moreover, in most countries the effect of mean school SES is stronger than an individual student's SES. (OECD, 2004).
- Rumberger & Palardy (2005) found that the effect of school composition was roughly equal to the effect of individual student SES in US secondary schools. They also found that the effect of mean school SES was similar for both high and low SES students.
- PISA study found that on average, school-level factors explain only about 5% of the between-school variation in student achievement in all the countries included. School level factors fall into three categories: school resources, school climate, and school policies.
- Much more important than school-level factors are student background and peer effects. On average, 50% of the between-school differences is due to individual student SES, 20% is due to mean SES of the school, 5% is due to school-level factors, and the remaining 30% is unexplained.
- The most important determining factor seems to be institutional arrangements and policies made at the national level rather than factors related to particular schools.
- Most of the countries that have the lowest effects of school composition have comprehensive secondary school system similar to that found in the US, Canada or Australia.
- In most countries student academic performance is greater in private schools. Once student SES is controlled for, however, private school advantage disappears or becomes minimal in most OECD countries. In other words, private schools have a performance advantage because of their student characteristics, not school-based differences.
- Overall, the social composition of schools can positively or negatively affect a family's choice of school. At the same time, school choice can positively but more often negatively affect a school's composition.
- Studies have found that low income students benefit more from educational interventions than higher income students.
- In four countries - Korea, Kenmark, Finland and Iceland- mean school SES has no statistically significant effect on student achievement in reading literacy (OECD, 2005).
- Recent studies form the US have started to document the positive effects of culturally diverse university settings for both minority and majority students. For example, interaction with culturally diverse students inside and outside the classroom is correlated with higher levels of complex thinking (Antonio et al., 2004), cultural awareness and political participation (Johnson & Lollar, 2002), and active thinking, tolerance and empathy (Gurin, Dey, Hurtado, & Gurin, 2002).
Outcomes, Racial Composition, SES Composition
Method of Analysis:
Unit of Analysis:
- Data comes from PISA (Programme for International Student Assessment) studies conducted by the Organization fro Economic Cooperation and Development (OECD).
- DV: Mean performance scores in reading, mathematics, and science.
- IV: Individual and school SES backgrounds (including means school SES).
- Rumberger & Palardy (2005)
- Henry and Rickman (2007)
- Lamb and Fullarton (2002)
- Robertson & Symons (2003)
- Rickman (2007)