The so-called “like school” comparisons in the My School website are a travesty. Apart from the flaws in the measure of the socio-economic status of schools, many background factors which influence school results are ignored. A key factor which is ignored is the ethnic profile of schools.
Children from particular ethnic groups are often concentrated in particular areas and schools, and children from some ethnic communities are highly concentrated in government schools. Performance disparities between “like” schools may reflect differences in ethnic composition rather than differences in school practices.
The Australian Curriculum, Assessment and Reporting Authority (ACARA) has stated that the ethnic background of students is excluded from the measure of “like schools” on My School because the correlations between the percentage of people who do not speak English well and student achievement were found to be below statistical significance [ACARA 2010].
This is not surprising because of the way in which ACARA has considered the impact of ethnic background on student achievement. It has considered these students as one group instead of consisting of different ethnic backgrounds that are differentially associated with levels of student achievement.
When all students from ethnic backgrounds, or from non-English speaking backgrounds, are considered as a group they achieve at similar levels to students of Australian-born parents [Cresswell 2004; Cobbold 2009]. There is little evidence that students with language backgrounds other than English have poorer educational outcomes. This has been the case for many years [Sturman 1997; Ainley et.al. 2000].
However, there are significant differences in student achievement between different groups of students from non-English speaking backgrounds. ACARA has mistakenly analysed only a very broad category of people and ignored differences within the group of people who do not speak English well.
Asian students, in particular, Chinese students, appear to achieve at higher levels than other groups. Students from some ethnic groups achieve much lower average results.
There is evidence that the average results of Lebanese students and Pacific Islander students are well below those of Chinese students. A study of Year 10 Certificate results for English, Mathematics and Science in three high schools in south-western Sydney found substantial differences between the results of students from different ethnic backgrounds [Suliman & McInerny 2006]. It found a much higher percentage of Lebanese students were achieving in the lower grades in all subject areas than Chinese and Vietnamese students.
It also found that a much lower percentage of Lebanese students achieved the top grades than Chinese and Vietnamese students. The average results of Lebanese students were significantly below those of Chinese and Vietnamese students.
An analysis of data from the Longitudinal Surveys of Australian Youth shows that achievement in literacy and numeracy in Year 9 varies widely among students from language backgrounds other than English [Marks & McMillan 2000]. When ethnic background was measured by nine categories relating to father’s country of birth, some ethnic groups showed higher Year 9 achievement levels than those students with fathers born in Australia, while students from other groups showed lower Year 9 achievement levels.
This finding was confirmed by further analysis of university entrance scores. It found that students classified as Asian performed substantially better than students whose fathers were born in Australia when adjusted for socio-economic background [Marks et.al. 2001]. On the other hand, the mean scores of students with Middle Eastern, North African and Pacific Islander ancestries were significantly below those of Asian students.
A study carried out by the Australian Council for Educational Research (ACER) for the National Education Performance Monitoring Taskforce of the Ministerial Council for Education, Employment and Youth Affairs (MCEETYA) shows that average numeracy test scores for Year 9 students from the Middle East, North Africa and the Pacific Islands are significantly below those of Asian students [Ainley et.al. 2000].
A later ACER report which analysed the 2000 PISA results for Australian immigrant children found that average reading and mathematics scores for Chinese and other Asian language students were significantly higher than for students with Middle Eastern home languages [Cresswell 2004]. In terms of proficiency levels, nearly 50% of Chinese students and 29% of Other Asian students achieved at the top two reading levels compared to 15% of Middle Eastern students. In contrast, only 9% of Chinese students did not achieve expected minimum standards compared to 24% of Middle Eastern students.
Such disparities in outcomes for students from different ethnic backgrounds are concealed by aggregate measures as used for the My School comparisons. The ACER report to the Performance Monitoring Task Force states:
….it needs to be acknowledged that there is a great diversity in the educational outcomes of students from ethnic minorities. Students from some ethnic or language backgrounds perform better than others and the use of aggregated data conceals these differences. Consequently, it is important to collect detailed data on the cultural and language backgrounds of students. 
It recommended that when using country of birth and language measures to report educational outcomes, students should be grouped into several categories [Ainley et.al. 2000: 32]. It is curious that this recommendation was not adopted by My School when one of the co-authors of the report is on the expert panel advising ACARA on the measurement of the socio-economic status of schools.
A recent report by ACER to the MCEETYA Expert Working Group on reporting for school evaluation and accountability also recommended that national reporting on school performance should include the language backgrounds of students in schools should be taken into consideration in any evaluation of school performance [Masters et.al. 2008: 24].
The failure of My School to take into account the different ethnic composition of schools has important implications for the accuracy and validity of like school comparisons. Performance disparities between so-called like schools may reflect differences in ethnic composition rather than differences in school practices.
For example, schools with a similar socio-economic status ranking on My School could have a high proportion of students of Asian origin while others have a high proportion of Middle Eastern or Pacific Islander students which accounts for the difference in average results. Parents and others using the like school comparisons of My School could therefore be misled about the comparative quality of teaching and curriculum in these schools.
This failure to properly take into account the impact of ethnic composition in comparing school results is yet another reason for a full public independent inquiry on My School.
Ainley, John; Frigo Tracey; Marks, Gary N.; McCormack, Silvia; McMillan Julie; Meiers, Marion and Zammit Susan A. 2000. The Measurement of Language Background, Culture and Ethnicity for the Reporting of Nationally Comparable Outcomes of Schooling. Report for the National Education Performance Monitoring Taskforce of the Ministerial Council for Education, Employment and Youth Affairs, Australian Council for Educational Research, August.
Australian Curriculum, Assessment and Reporting Authority 2010. Index of Community Socio-Educational Advantage (ICSEA), My School Technical Paper.
Cobbold, Trevor 2009. The Free Market and the Social Divide in Education. Dissent 31: 12-19.
Cresswell, John 2004. Immigrant Status and Home Language Background: Implications for Australian Student Performance in PISA 2000. Australian Council for Educational Research, Camberwell, October.
Marks, Gary and McMillan, Julie 2000. Social Background and Educational Outcomes: Preliminary Results from the Longitudinal Surveys of Australian Youth. Report for the National Education Performance Monitoring Taskforce, June.
Marks, Gary; McMillan, Julie and Hillman, Kylie 2001. Tertiary Entrance Performance: The Role of Student Background and School Factors. Longitudinal Surveys of Australian Youth Research Report No. 22, Australian Council of Educational Research, November.
Masters, Geoff; Rowley, Glenn; Ainley, John and Khoo, Siek Toon 2008. Reporting and Comparing School Performances. Australian Council of Educational Research, December.
Sturman, Andrew 1997. Social Justice in Education. Australian Education Review No. 40, Australian Council of Educational Research, Camberwell.
Suliman, Rosemary and McInerny, Dennis 2006. Motivational Goals and Student Achievement: Lebanese-background Students in South-Western Sydney. Australian Journal of Education 50 (3): 242-264.