Standardized testing

October 30, 2018

“Our goal is to prepare Ontario students for success, improve their academic achievement and equip them with the tools needed to enter the working world.”

Ministry of Education, October 2018

Ontario currently tests every student in grades 3 and 6 in reading, writing and mathematics; in grade 9 in mathematics, and in grade 10 in literacy. Most other provinces in Canada have similar testing regimes.

While the public, the media, and many school boards find the test results useful, a number of policy experts, academics, and research-based organizations (People for Education among them), have raised concerns about the consequences of focusing system goals too narrowly on the results of tests in three subjects.

When test scores are used as a proxy for overall success of the system, it can lead governments to target funding and policy in ways that can constrain risk-taking and experimentation, shift resources away from other areas and ignore vital skills and competencies that are foundational for students’ long-term success. For more information, see The New Basics, People for Education’s 2018 Annual Report on Ontario’s Publicly Funded Schools.

Key principles for large-scale assessment

People for Education – working with experts from across Canada and educators in Ontario schools – has been leading an initiative to identify, define, and test essential competencies for long-term success.

We believe that any revisions to the province’s standardized testing policy should be based on the following principles:

To be effective, and to ensure coherent education policy, standardized testing strategies must be integrated within an overall framework for education.
- Large-scale assessment is only one piece of a larger education system, and should not be examined in isolation from other aspects of education reform.
- It is critical to recognize that the most productive assessment – the assessment that leads to the deepest learning and development, and has the greatest impact on both students’ success and teachers’ practice – happens inside classrooms.
Information about success must go beyond test scores in three areas.
- There is wide-spread agreement that public education has multiple purposes beyond teaching literacy and numeracy skills. The public education system plays a central role in ensuring that the next generation is equipped to address the complex social, economic, and environmental challenges that the world is facing. To ensure their long-term success – both economic and social – students need to develop foundational citizenship, social-emotional, health, and creative competencies, including things like the capacity to innovate, think critically, collaborate, solve complex problems, transfer knowledge from one area to another, set goals, deal with failure, and understand others’ perspectives.
- It is difficult, if not impossible, for standardized tests to assess these vital and complex areas of learning. However, information collected through things like student, teacher and principal surveys, can provide vital information – both to the public and to policy-makers.
There is danger in confusing the proxy for the whole pie.
- Large-scale assessments can create a false assumption that achievement on a single test provides a complete picture of a student’s ability in that area. It is critical to acknowledge that standardized assessments or surveys can only give limited information about complex competencies
- Just as achievement on a single test does not give a complete picture of an individual student’s ability, provincial results on standardized tests are limited as a proxy measure of overall system success.

Sampling versus census-based testing

One way to realize the benefits of large-scale assessments, while mitigating the challenges, is through sampling, or randomized testing. Just as polling companies survey representative samples of the population to identify social trends and opinions, and PISA tests conducted by the OECD use results from a random samples of 15 year-olds (In Canada’s case, just 20,000 students from across the country) to report on a whole country’s success in reading, writing, mathematics and science, provinces can test a sample of students to get information on students’ overall acquisition of knowledge in a range of areas.

The use of sampling would make it difficult, if not impossible, to use standardized test results to rank schools and school boards, which was never the intended purpose of testing in Ontario.

Moving away from whole population testing creates some challenges:

Currently, school boards may use these achievement indicators to target resources and support to schools, so that students have more equitable outcomes. Sampled testing might make it more difficult for boards to identify these schools.
It may also make it slightly more complex to compare schools with similar demographics – one of the methods the province, school boards and educators currently use to evaluate policy and strategically direct resources. It may also make it more slightly more difficult for school boards to report about student performance to their local communities.

However, these challenges are not insurmountable.

Sampling student results, rather than testing every student, might allow the province to evaluate a broader number of areas – perhaps on a rotating basis, as PISA does.
By using sampling, rather than testing every student, it may be easier to report to the public on a broader view of educational success in the province. That larger picture could include things like findings from the Early Development Instrument, the EQAO’s student, teacher and principal questionnaires, school climate surveys, or new surveys such as the British Columbia Student Learning Survey, conducted annually to grade 4, 7, 10, and 12 students in BC.
Standardized assessments given to a sampled population of students, on a sampled selection of competencies, could provide information about system performance, while avoiding some of the negative consequences of large-scale measurement.
By collecting demographic information from the sample of students who are tested, the province could still analyze the results to identify gaps in achievement between groups based on socio-economic status, race, special education needs, and many other factors. This means that resources and support can still be targeted to the groups of students who need it most.
As an altnerative to standardized tests, or in addition to sample tests, regionally-based teams of teachers could develop standards for a range of areas of learning, and conduct team-based assessment at different levels/grades in the system. This would provide jurisdictional information about student performance, but would require more resources, infrastructure, and capacity.