Part 1: The role of socio-education advantage and HSC performance
It is conventional wisdom that students from more wealthy and advantaged families do better at school. Each year when the NSW HSC results are released, the top rankings are dominated by selective schools such as James Ruse and expensive private schools such Sydney Grammar School. In fact, in 2019 the top performing non-selective government school came in at 67 (Willoughby Girls High) and the top systemic Catholic school just ahead at 64 (Brigidine College, Randwick).
The chart above illustrates clear groups between school performance and tuition fees for each of the four main school sectors: Independents, Government selective, Government non-selective, Catholic system (different from independent Catholic faith schools).
A simple glance at the chart shows that if you want your child to be in a top 60 ranked school then they either need to get into the best selective schools or you’ll need to spend at least $18,000 pa (and increasing by 3–5% each year).
So what makes these schools so good?
ICSEA: Accounting for Educational Advantage
There are many factors that affect different outcomes in our lives. For example people with higher levels of education tend to earn more and live longer. When determining whether an initiative is successful, it is important to account for these other factors to see what is causing the change.
The Index of Community Socio-Educational Advantage (ICSEA) is a measure of the educational advantage of a school’s student cohort. ICSEA scores allow better comparison between schools by taking into account the relative advantage of each school’s student body.
ICSEA takes into account factors such as parental occupations, educational levels and geographic location. The average (median) school score is 1000 and the standard deviation is 100. Accordingly, schools with scores above 1100 are considered quite advantaged while schools with scores below 900 are quite disadvantaged. ICSEA scores do not directly account for family wealth or income, although it is likely indirectly captured through occupation status.
Further information can be found in my previous blog post.
ICSEA explains much of the variation in NAPLAN scores
The reason why ICSEA has been used prominently is that it does a good job (from a statistical basis) in predicting academic performance. It is not a perfect measure and there are numerous issues with NAPLAN however the data linking ICSEA with NAPLAN performance is robust.
ACARA’s 2017 Technical report on ICSEA found that 80% of the variance in school NAPLAN performance can be explained by differences in ICSEA scores. This is consistent with the 2015 report which found that 78% of the variation was explained by ICSEA.
In layman's terms this means that the more socio-educational advantage a student has, the more likely that student is to perform strongly.
ICSEA also correlates well with HSC performance in government schools…
The data above showed a clear link between ICSEA scores and NAPLAN performance. A common critique of NAPLAN is that it only tests a small subset of the curriculum and can be gamed by teachers who teach to the test.
In contrast, the later years of school offer a broader subject range. Given this, it is initially less clear that ICSEA would correlate well with HSC performance. This is something that I wanted to test.
Each year in December the HSC results are released along with media articles devoted to school rankings. There are many ways in which you could rank schools however the most common is based on the proportion of students who score a Band 6 in the subject, the highest proficiency level.
I compared the HSC performance of schools, according to the percentage of Band 6s achieved, against the ICSEA scores (MySchool website). HSC performance was sourced from Better Education, and ICSEA scores were sourced from the MySchool website.
The chart below shows the results for the top 500 performing schools in NSW. From the graph it is clear that there is a significant correlation between ICSEA scores and school HSC performance. While the analysis is rough it does suggest that a large portion (~77%) of the performance difference between schools is explained by educational advantage — schools with students from higher socio-educational backgrounds perform better.
As you can see on the chart above the line of best fit is non-linear (curved). Two potential reasons are
- Achieving band 6s across subjects are correlated: Students who get a band 6 in one subject are much more likely to get a band 6 in a few others. A crude way to say this is that (academically) strong students tend to be (academically) strong in many subjects. For example imagine two students who both do Maths and Physics. You are more likely to get a band 6 in Physics if you also got a band 6 in Maths than if you got a band 4 in Maths.
- Advantage accumulates, namely better schools attract better students: High potential students who can go to high performing schools on average will. This has the effect of concentrating performance in a small set of schools and conversely hollowing up the remaining schools.
In the following sections we investigate the relationship between ICSEA and HSC performance across the different school sectors.
ICSEA also correlates well with HSC performance in government schools…
The chart below looks at just government schools. Red dots represent selective schools and green dots represent comprehensive or non-selective public schools.
The chart shows that the best performing government schools (all selective) also have the highest ICSEA values. For example the minimum ICSEA score for schools achieving more than 30% band 6s is 1100.
The reverse also appears true (although you would need individual level data to know) — high ICSEA families, whose children are at public schools, are choosing to send their children to selective schools over comprehensive public schools. Government schools with ICSEA scores above 1150 (1.5 standard deviations from the mean) are all selective.
This leads me to the following question:
Are the top selective schools actually better or are they just picking better students?
I wonder whether those students at selective schools would have performed similarly well if they had attended the local comprehensive high school?
Let’s now turn our attention to Catholic and Independent school sectors
The correlation between ICSEA and academic performance is weaker in Catholic and Independent schools
When looking at the Non-government schools the correlation between HSC performance and ICSEA is not as strong. Yet ICSEA still explains about 57% of the variation in school HSC performance.
As before we can split the data between Catholic and Independent schools. In general the higher performing schools are Independent rather than Catholic with the top performing Catholic school achieving 23% of scores as Band 6s.
A few interesting points to note from the data
- There is a larger range of ICSEA scores within the Independent sector (949–1277) than the range of scores in the Catholic sector (956–1123).
- The top performing Catholic system school scores less than 25% of band 6s. In contrast there are many very strong performing Independent schools. In general they are more advantaged i.e. higher ICSEA.
- While high ICSEA scores (above 1150) do appear to correlate with higher Band 6s, there are a number of schools that seem to underperform vs expectations i.e. attaining less than 20% Band 6s.
- For schools with ICSEA scores between 1000 and 1100 neither sector clearly outperforms the other.
For non-government schools, ICSEA is a better predictor of fees than HSC performance
In the chart below I combined Catholic and Independent schools and compared how ICSEA compared with both Band 6s and the schools fees charged.
It turns out that for Independent school ICSEA scores correlate more with fees than with performance. Schools are better at turning educational advantage into money than into results.
The link between educational advantage and fees makes sense. Families with higher educational advantage have more education themselves and therefore are more likely to have higher household incomes and a higher capacity to pay.
This relationship is not the same with Catholic schools.
Summary — ICSEA has a large impact on how schools perform in the HSC
In the above blog I have shown that ICSEA scores can explain a large part of HSC performance difference between school sectors and individual schools.
Schools with higher levels of educational advantage will on average perform better in the HSC.
Given the school ICSEA scores are derived from family ICSEA scores this implies that if you come from a high ICSEA family then you are much more likely to perform well academically. This effect is most pronounced in selective schools and least observed in Catholic system schools.
So how does this help you as a parent making a decision about high school for your child?
Thus far all I have demonstrated is that students from higher socio-educational backgrounds are more likely to perform better on the HSC. By the time your child is entering high school this has already been determine since ICSEA depends largely on parental education and occupations.
To make a decision on the academic potential of school you really need to understand how good a school is at value-adding their students. Any school should be able to do well if they enroll academically strong students. Like great coaches, great schools will help their students perform better than they otherwise would have.
In the second part of the blog series I will use these ICSEA data to show which schools are underperforming/overperforming against ICSEA expectations.
Jesse Whelan is the Founder of Sandbox Learning Australia an academic support organisation that helps students who have fallen through the cracks at school gain confidence, learn and succeed so they can thrive now and in the future.
He is passionate about making high quality learning accessible for all and is working on creating a low-cost private school concept. Reach out if you would like to be part of this new innovation.