International Schools' Assessment

Logo of ISA

FAQ - General Questions

When you say this is a test for International Schools, what do you mean by that?

Because the ISA is designed for students experiencing many different curricula, the assessment is geared to the skills and understandings that research suggests underpin major (western) curricula throughout the world. The frameworks for reading and mathematical literacy are based on those used for the OECD's Programme for International Student Assessment (PISA) which has been developed and endorsed by all OECD countries and is currently used by close to 60 countries to measure and compare student performance across countries. View Explanatory Materials/Reports Page for more details about the development and construct for the ISA

What distinguishes the ISA from other standardised assessments currently used in international schools?

It is designed specifically for international school populations

This means that it is a test that is not targeted to any one national or cultural group: the test material is eclectic, drawing on many cultural and national sources; it has been designed with the knowledge that the students who will sit the test come from many cultural, social and linguistic groups, and that the curricula they have been exposed to are diverse.

Another related feature is that the language in the assessments is chosen with an international population in mind, in the knowledge that at least half of the test-takers have mother tongues other than English. Although the reading and writing tests inevitably depend on language proficiency (inseparable from the domains), the mathematical test attempts to use language that will be accessible to non-English speaking background students as well as English speaking background students.

It is based on an internationally endorsed concept and assessment frameworks

The ISA is based on the construct and frameworks of the Organisation of Economic Cooperation and Development's Programme for International Student Assessment (OECD PISA). PISA was developed as a measure of 15-year-old student performance in reading, mathematical and scientific literacy. The PISA test concepts were created by teams of international experts in these fields. The ISA has taken those concepts and applied them to parallel and younger age groups. Some PISA tasks have been used in the ISA so that the ISA can be linked statistically to the PISA results. Thus ISA results are comparable with results from the PISA study

It measures growth over time for both individuals and school populations

Many standardised tests provide normative information and may also offer criterion-based reports. Because of the methodology used to analyse the assessment data for the ISA, we are also able to construct scales that are stable over time, and across year levels. This means that the results reported for students in Grade 3 are comparable with the results reported for the same students when they reach Grade 5 and thus the learning growth of the cohort can be monitored. Similarly, one year's Grade 3 results are comparable with next year's Grade 3 results, so that a school can monitor program development at a particular grade level and across the school.

Is the test on a Southern Hemisphere or Northern Hemisphere school cycle?

The timing of the ISA was originally set in relation to the northern hemisphere academic year which most international schools follow. Almost all students taking the October ISA, therefore, have only had a few weeks in the designated grade when they take the assessments. The reason for conducting the testing early in the school year is so that the information reported can be used by schools, and particularly classroom teachers, to inform their instructional programs within the same academic year. It is thus strongly formative in its intentions. Note that there is a three-month turnaround between test administration and issuing of results.

That said, in response to feedback from schools, since ISA 2005-2006 we have offered schools the option of administering the ISA in either October or February. For students continuing into the following year, their results from the February assessment could be used to inform their new teacher about their stage of development at the beginning of the new academic year.

The first February administration occurred in 2006 and we will continue to offer this option in future years.

How many schools participated in the previous administration of the ISA?

107 schools took the ISA in October 2007, with a further 111 schools testing in February 2008. 23 schools took part in the ISA trial test in late January 2008.

In October 2007 over 21,000 students from 107 international schools in 50 countries participated in the ISA. 97 of the 107 participating schools run on a Northern Hemisphere academic calendar. The numbers by grade level were 3617 at Grade 3, 1605 at Grade 4, 4181 at Grade 5, 1958 at Grade 6, 4062 at Grade 7, 1988 at Grade 8 and 3638 at Grades 9 and 10 (1625 and 2013 respectively). About 41% of the students were from English speaking backgrounds, and 59% from non-English speaking backgrounds. In October 2007 slightly more boys than girls took the assessments, with boys comprising between 51% and 53% of each grade level. In February 2008 over 16,000 students participated in the ISA. View Participant Schools

Can we test in both October AND February, which should we choose?

Schools have the option of administering the ISA in EITHER October OR February. Our policy is not to allow testing at the same school on different occasions, since the material is identical in October and February. This avoids test security being compromised with the chance of testing material being leaked. The ability to compare results over time would also be adversely affected.

If your school's students are continuing into the following year, then the results from the February assessment can be used to inform their new teacher about their stage of development at the beginning of the new academic year. The October administration - for which results are provided at the end of the following January - would provide the teacher with feedback about his or her students within the same academic year - though, half-way through.

What is the effect of moving from the October Administration to the February Administration of the ISA, or vice versa?

Changing the timing of a school's testing may impact upon trend monitoring. Overall, we see that there is a significant difference in performance for Grade 3 between October and February. The difference declines as students get older, so by Grade 10 on average there is virtually no effect. However, at an individual school level there may be quite a different picture, so any interpretation of trend results would be dubious. If a school changes the testing time, trend collection should start again.

Is the test designed for Australian students?

The ISA is designed specifically for students in international schools worldwide. A few internationally-minded schools in Australia have taken the ISA each year: in 2007-8, for example, of the 218 schools that participated, only two were located in Australia.

Although there are some Australian-developed educational theories embedded in the ISA, such as developmental assessment, and the test is written and analysed in Australia, the ISA is emphatically an assessment conceived on international principles, based on the frameworks and scales of the OECD's Programme for International Student Assessment (PISA) that is developed and overseen by international experts in mathematics and language education.

Is the ISA compatible with the IB inquiry-based approach?

We have conducted an audit of the PYP and MYP and in general our finding was that many educational principles are common to the IB programs and the ISA, notably the emphases on empowering and encouraging students to become life-long learners, on reflective thinking and on responsibility towards oneself, the community and the wider world. More narrowly, the basic principles of assessment enunciated by the primary and middle years IB programs and by the ISA are also in accord, with priority given to meaningful reporting to parents, the provision of useful information about the needs of individual students, and the provision of feedback to improve teaching and learning.

Which grades can do the ISA?

From 2002 to 2005, the ISA was available for Grades 3, 5, 7 and 9/10. Since October 2006 the ISA has been available for Grades 3, 4, 5, 6, 7, 8 and 9/10.  Schools have the option of registering their students from any or all of these grades. Please note that the test is available to students in either Grade 9 OR Grade 10, not both.

Do you use the same test for Grade 9 and Grade 10?

Yes

Can we test our students in Grade 9 and in Grade 10?

No, a school needs to choose which Grade they prefer to test. Additionally, schools cannot test Grade 9 students one year and then register Grade 10 students the following year as the same students would be sitting the ISA Grade 9/10 test (firstly as Grade 9 students, then as Grade 10 students in the following year). While the tests for Grade 9/10 in consecutive years are not identical, there is a significant overlap of material, and our research evidence suggests that there is a concomitant significant effect on results if students in this age group are administered the same test material twice. From the school's perspective, this is likely to mean that data would not be usefully comparable in future years. We are also concerned to protect the validity of the results for all schools.

Why can students in Grade 9 or Grade 10 take the same test?

We are able to give the same assessment to Grade 9 and 10 because we can target the range of abilities typical of those grades with the same set of tasks. However, we report Grade 9 and Grade 10 separately, using different comparison groups depending on the grade to which your school has administered the assessment.

There is typically a difference in performance between grades 9 and 10 but it is smaller than the difference between grades 3 and 4, for example: in fact the older the students, the smaller the difference. The narrow difference between Grade 9 and 10 may not be representative of all Grade 9s and Grade 10s; it depends on the sample of schools that takes the ISA in any one year. The general picture that emerges from our data set is, nevertheless, consistent with what we see in other measures and assessment programs. Indeed, in some jurisdictions performance in two consecutive grades actually can go down from one year to the next (especially where there are transitional issues, such as changing from elementary to high school).

What criteria would you suggest we use to determine a Grade 9 or Grade 10 testing year?

The same test is administered to Grade 9 and Grade 10 students. Whether a school decides to administer the test to Grade 9 or Grade 10 can be determined by a number of factors. For example: if a school has an external examination like the IGCSE for Grade 10 students, it may not wish to give another external assessment in the same year, and so would choose a Grade 9 administration. On the other hand, if a school is strongly focused on comparing its results with national norms of 15-year-olds from PISA, and the average age of students in Grade 10 at the school is 15, then a school might choose the Grade 10 administration. Again, if the school wants a benchmark measure of the final year before (say) the IB diploma program starts, Grade 10 would be the better choice.

Is it possible to do just one subject?

In the ISA, we assume that students will take all four parts of the test (Mathematical literacy, reading and two writing tasks), and we charge on that basis. A school may elect to do one, two, three or four parts of the test but the charge is the same, regardless. View Costs

Are there test items that are the same in each of the grade levels tested?

Yes, we use common items between adjacent grade levels in any one year, as well as items that are common to consecutive years of administration (e.g. the same item may appear in Grade 3 2008 and Grade 3 2009). Both of these kinds of 'link' items are used to establish the scale, vertically (between grade levels) and longitudinally (over time). While we have some new items every year, about 1/3 to 1/2 of the items in any one form may be used as links either vertically or longitudinally.

How do you take account of the student's age in the ISA?

The ISA testing cohort is based on the student's grade, rather than on the student's age. In any survey across countries, comparing student cohorts is a complex issue. Some studies choose an age-based criterion and others a grade-based criterion, and still others a 'number of years at school' criterion. There are advantages and disadvantages of using each method. For the ISA we have chosen a grade-based criterion. We are using UNESCO's International Standard Classification of Education (ISCED-97) classification of grades. What we call Grade 3 / third grade / Year 3 is the third year of ISCED level 1; what we call grade 7 / seventh grade / Year 7 is the seventh year, counting from the beginning of ISCED level 1. Please note that the grade levels are not based on Australian year levels or the Australian school year: indeed, the great majority of schools participating in the ISA follow the Northern Hemisphere academic year.

We ask schools to submit the 'named grade' for the assessments: that is, children in the year or grade called 'three' should do the Grade 3 test, regardless of their age; children in the year or grade called 'five' should do the grade 5 test. The exception to this is children in British-style schools. The AGE of children in British style schools is approximately the same, on average, as that of children in the grade 'below' them. In the light of this, we recommend that children in British-style schools are administered the assessment of the Grade name BELOW their Year level. That is, Year 4 children should sit the Grade 3 tests; Year 5 children should sit the Grade 4 test etc.

Our analyses show that the age of the students is correlated with performance for the youngest Grade level, but that age decreases in relevance as the students become older. What is more important is the number of years at school. By Grade 7 there is negligible difference in terms of the age of students.

How much time has gone into ensuring that there are as few culturally specific references as possible?

The ISA has been developed with the needs of students from very diverse cultural backgrounds in mind. We draw on our considerable experience in catering for multicultural populations within Australia, and more recently, on our experience in developing assessment instruments for students from about 40 countries in the OECD's Programme for International Student Assessment (PISA). As with PISA, ISA's philosophy in dealing with the issue of diversity is to present culturally diverse material in our assessments, rather than to attempt cultural neutrality.

Regarding the issue of bilingualism: Although the ISA is being offered in English language only, we are very much aware of the fact that a large proportion of international school students are from non-English-speaking backgrounds, and are mindful of minimising the language loading of assessments in which language competence is NOT the focus of assessment (e.g. mathematical literacy). In the writing assessment, in which language competence is the focus of assessment, at least some of the criteria used to measure student performance will be designed with non-English speaking background students in mind.

And in reporting all student results to schools we will ensure that the language-background of students is acknowledged. Thus non-English speaking language background should not be a 'disadvantage', but level of language development should be fairly and validly measured.

Does the ISA test in the area of science?

There are no plans in the short term at least to include a Science test within the ISA: we are concerned to consolidate in the domains that we are offering currently: mathematical literacy, reading and two kinds of writing. Looking ahead, however, with Scientific Literacy as one of the major PISA domains, we will consider mounting an ISA science test if there is sufficient interest from participating schools.

How much assessment is recommended for students grade 3-10 as they progress through the school system?

The frequency and range of ISA tests that your school might administer depends on your purpose in using the test. For example, if you wish to check how you are going relative to others every so often you may elect to participate every second or even third year. It is unlikely that teachers will learn to benefit from the diagnostic information they can get from the ISA if you do this. If you wish to track individuals over time and you have a highly mobile population you probably need to test every year to capture most students for at least two assessment periods. If you use the ISA regularly, teachers are more likely to learn to use the diagnostic information constructively. It depends what your school needs from the ISA.