Manitoba students’ test scores and the scientific method

“Science is … properly understood as a framework for creating and furthering knowledge about how the world works…” (Sowmya Dakshinamurti, Free Press letters, Nov. 21, 2025). The just-published report on Grade 8 student achievement, focused on science in this round, provides an opportunity to put some science — observation, interpretation, knowledge claims, theories, prediction, repeat — to work.

The Pan-Canadian Assessment Program (PCAP) is a periodic, comparative assessment across provinces of Grade 8 students in science, reading and mathematics. PCAP 2023 joins years of corroborating results letting us claim as knowledge that the average scores reflecting Manitoba students’ academic achievement are among those of the lower-scoring provinces.

This knowledge is marred, however, by questionable, overdramatic interpretations.

Manitoba’s score differences relative to most other provinces range in significance from statistical noise to small or modest in practical terms of knowledge and skill. Based on the Organisation for Economic Co-operation and Development’s (OECD) Programme for International Student Assessment (PISA) targeting students at around Grade 10 (student sampling is age-based), Manitoba’s scores are at or a little above the average of those (38, currently) member countries.

We know, then, that Manitoba students’ average achievement is more similar to than different from that of other provinces; that it is a little below the Canadian average and is on par with, or a little better than, the average of OECD countries, most of which are developed nations.

Nevertheless, reactions often include wording such as “dead last” (inapplicable in this round) and “failing grade,” and our education system as systemically floundering, be it related to curricula, teaching emphases and methods, or provincial testing. Does a theory that results can be explained and notably affected this way stand up?

Changes to the teaching of science, mathematics and reading in the last 15 years — e.g., curricula, emphasis, strategies — are unaccompanied by notable changes to Manitoba’s rankings. This does not mean that such initiatives are without effect, be it negative or positive, just that they have no overpowering influence on comparative measures of learning like PCAP.

In the first-ever release from PCAP — PCAP 2007 — Manitoba students were in the middle of the rankings, as usual to that point, in reading and mathematics. In science the rank was eighth among provinces, a bit jarring as well as, it turned out, a harbinger.

A little over a year after the February 2010 release of that report came the science education action plan, with investments — $25 million over five years — in science classrooms and labs, professional development, curricular revisions and partnerships. Again, the rankings have remained about the same since.

As before, this is not to dismiss such initiatives. Rather, it further suggests that a theory of student learning must include factors that are more influential and outside of classrooms.

Socio-economic status, for example, is strongly associated with group-level measures of learning. The Social Planning Council of Winnipeg report Manitoba Child and Family Poverty (February 2024) provides a perspective — namely, high rates of child poverty — for the learning context in Manitoba.

This aligns with other Manitoba observations — crime rates, life expectancy, physicians per capita, equalization payments doubling over the past six years and at the highest proportion of federal payouts recorded in Government of Canada Historical Transfer Tables, and reflected even in a low rate of patent applications. (Positively, we are relatively generous according to the Fraser Institutes’ report Generosity In Canada: The 2025 Generosity Index).

Holding the education system primarily responsible for delivering what would be, clearly, anomalously elevated test scores, is unscientific.

Interestingly, PISA scores (for 15- and 16-year-olds) are on a downward trend across most provinces and most OECD countries, suggesting again other negative, broad-based influences on student achievement than those based in the education system.

Certainly, let’s try to improve educational practice, and use assessment data to assist. With other issues properly dominating headlines, and without the advocates and activism associated with mathematics and reading, perhaps this report will evade much attention.

If there is any, the next science test will be to see if the resulting rhetoric, expectations, efforts and investments for improving student achievement are guided by “how the world works.”

Ken Clark spent most of his time while in the field of education specializing in large-scale assessment, like PCAP, and assessment policy.