The EQAO Grade 9 math assessment: what it is, how it is scored and reported, whether it counts, and how to prepare in three weeks

The EQAO Grade 9 math assessment is a provincial test written by students in Ontario's Grade 9 math course, set and reported by the Education Quality and Accountability Office. It is written online, in your school, and your result comes back as an achievement level rather than a percentage. It counts toward your course mark, by a share your school or board sets inside the Ministry's range. Here is what it tests, how it is reported, and a three-week plan that uses the agency's own sample test.

What the EQAO Grade 9 math assessment is

The agency is Ontario's arm's-length office for provincial testing. It sets the test, scores it and publishes the results for the province, each board and each school. Your teacher does not write it and does not mark it.

Who writes it: students taking the Grade 9 math course, code MTH1W. That is the de-streamed course that replaced the old academic and applied Grade 9 courses in 2021, so almost every Grade 9 student in a publicly funded school writes the same assessment. When: during the semester you take the course; the school picks the days inside the agency's window. How: on a computer, in two sessions, each built to take about an hour. A session has two stages, and the questions in the second stage are chosen from how you did in the first; the agency calls this a multi-stage adaptive design. Your school sets the days.

The purpose, in one sentence: the province measures how the system is doing, and you get a level. The long-run reason to care about it is simpler than that. It is a review of the whole Grade 9 course, and the Grade 9 course is the floor under every math course after it, through to the Grade 12 courses that university programs ask for.

What it tests: the course's areas, in our words

The assessment covers the five content areas of the course. In our own words, not the Ministry's:

• Number: fractions, integers, powers, ratios, rates and percentages, and the arithmetic that holds them together.
• Algebra: writing and simplifying expressions, solving linear equations, and linear relations with their tables, graphs and equations.
• Data: collecting data, reading tables and graphs, and the measures of centre such as the mean and the median.
• Geometry and measurement: perimeter, area and volume, and the relationships between angles in lines and shapes.
• Financial literacy: interest, budgets and comparing the cost of choices.

Every question is answered on screen. The agency's own page lists the kinds: pick one answer, pick more than one, drag items into place, and choose from a drop-down menu. None of them is a long written solution. The whole thing is scored by computer.

That has one consequence for how you practise: the agency's sample test is the only practice test worth doing, because it is the real format. A worksheet from somewhere else may be good math practice; it is not EQAO practice.

How it is scored and reported

The agency does not turn your raw score into a percentage. It uses a statistical scoring method to place you on a scale, and the scale is reported as one of four achievement levels. Level 3 is the provincial standard, the level the province expects. Level 4 is above it. Levels 1 and 2 are below it.

Your school receives your individual student report, showing your level, shortly after you finish. The results for the province, your board and your school are published later on the agency's site.

What a level is not: it is not your percentage on the test. It is not a rank among your classmates. And it is not a permanent label. It says where you were on a day in Grade 9. A Level 2 names the areas to work on in Grade 10, not the student.

Does it count?

Yes, and the rule is written down. The Ministry of Education's assessment policy says the result of the Grade 9 assessment is part of the final evaluation of the Grade 9 math course. The weight is at least 10 per cent and at most 30 per cent of the final mark, and your school or board sets the share inside that range; several boards use 10 per cent. Your teacher will tell you how that works in your class, because the agency sends the school your result and the school builds it into the mark.

What to do with that: ask the teacher in the first week how the assessment counts, write the answer at the top of your plan, and prepare as if it matters. It does, and the preparation costs you nothing extra, because preparing for it is the same thing as reviewing the course.

The method: one sample test, an error log, then the plan

This is the practice-test method and the error log from the exam guides, cut down for a fourteen-year-old with a laptop and three weeks.

1. Take the agency's sample test under real conditions. On a computer, in one sitting, with the on-screen tools and nothing else: no notes, no phone, no help. It is one session with two stages, shorter than the real thing, and that is fine; what you are measuring is you.
2. Mark it. Where the sample does not show you an answer, check it with your teacher or redo the question slowly with your notes open. Every question you got wrong, or guessed, is one row in an error log: the area it belongs to, the kind of error (did not know how; knew how but slipped; ran out of time), and the fix in one line.
3. Rank the areas by the number of rows. The top three are your three weeks. The rest get a light touch.
4. Drill the top three with mixed sets and recall, not re-reading. Rohrer and Taylor (2007) gave students the same practice problems either grouped by type or shuffled, and the shuffled students did worse in practice and better on the test a week later, because they had practised deciding which method a question needs. That is exactly what the real assessment asks. Mixed sets shows how to shuffle your own practice. And test yourself before you look anything up: self-testing, not re-reading, is the single method with the strongest evidence behind it.
5. Retake the sample a week before. Roediger and Karpicke (2006) found that taking a test is itself one of the strongest ways to remember, so the second run is practice as well as a measurement.

Worked example: three questions of our own, with the wrong line

These are our questions, not the agency's. Each one shows the kind of error the log would catch.

Algebra: a linear relation. A gym charges 20 dollars to join and 5 dollars for each visit. What do 12 visits cost?

• Cost = 20 + 5 × 12.
• 5 × 12 = 60.
• 20 + 60 = 80. The answer is 80 dollars.

The wrong line: (20 + 5) × 12 = 25 × 12 = 300. The 20-dollar fee is paid once, not every visit, so it does not get multiplied. In the log this is "did not know how", and the fix is "the fee you pay once is the constant; only the per-visit cost is multiplied".

Data: mean and median. Five test marks: 72, 85, 91, 68, 84.

• The mean: 72 + 85 + 91 + 68 + 84 = 400, and 400 ÷ 5 = 80.
• The median: sort first, 68, 72, 84, 85, 91; the middle value is 84.

The wrong line: taking the middle of the unsorted list, 91. The median is the middle of the sorted data. In the log this is usually "slipped", and the fix is one word: sort.

Measurement: area of a path. A rectangular garden is 6 m by 4 m, and a path 1 m wide runs all the way around it. What is the area of the path?

• The garden plus the path is a bigger rectangle: 6 + 1 + 1 = 8 m by 4 + 1 + 1 = 6 m.
• The big rectangle's area: 8 × 6 = 48 m².
• The garden's area: 6 × 4 = 24 m².
• The path: 48 − 24 = 24 m².

The wrong line: 7 × 5 − 24 = 35 − 24 = 11 m², adding the path's width once instead of on both sides. A sketch catches it in ten seconds. In the log the fix is "a border adds its width to both ends of each side".

Three answers to check your own against: 80; 80 and 84; 24.

The three-week plan, on paper

Three weeks before the assessment, a table you can copy onto one page. Each session is about 25 minutes. "Recall" means a blank page with the idea's name at the top and the notes closed; "mixed set" means a handful of questions from more than one area, shuffled.

Week	Sessions	The recall task	The mixed set	Minutes
1	Day 1: the sample test, timed, on a computer; mark it; write the error log; name the top three areas	None yet	None yet	60 + 30
1	Three more sessions: for each of the top three areas, one worked example from the class notes read slowly, then redone cold	The method of the example, written from memory	Three questions from that area	25 each
2	Every second day, four sessions	One idea from each top area, recalled cold	Six questions across all five areas, shuffled	25 each
2	One session on the format: the sample test's question types and on-screen tools, so the format costs nothing on the day	None	None	20
3	Five days out: the sample test again, timed; a new, shorter error log	None	None	60 + 20
3	Two sessions on the two weakest areas from the new log only	Those two ideas, cold	Six questions, those areas mixed with one easy area	25 each
3	The last two days: light. One short recall each evening, then stop. A full night's sleep before the day	One idea	None	10

Two habits for the whole three weeks. First, the open-response habit even though the test is on screen: on paper, write the step, not only the number, because the step is what you can check and the step is what goes in the log. Second, worked examples are read slowly and then redone with the example covered. Sweller and Cooper (1985) found that students who studied worked examples before trying problems learned algebra procedures faster and made fewer errors; how to study math has the full cycle. If the nerves arrive in week 3, the page on test nerves has the two-minute version, and chapter 9 of the free playbook is the seven-day exam plan the last row is cut from.

For parents: what helps and what does not

What helps. The plan above, on the fridge, with the sessions ticked off. Reading the error log together once a week and asking "what did you get wrong, and why" rather than "what did you get". Protecting the sleep on the two nights before. Treating the sample test as a measurement, not a verdict.

What does not help. A practice test of unknown origin, which may be the wrong format, the wrong difficulty, or simply have wrong answers. Extra worksheets on areas the error log does not name. Hiring help for "Grade 9 math" in general when the log says "fractions and the median". Treating a level as a judgement of your child.

One sentence to keep. A Level 1 or 2 is the agency's way of saying which areas need work in Grade 10; it is information, and the student who writes the error log already has more of it than the report gives.

The assessment, on one screen

• What it is: a provincial online test in the Grade 9 math course, written in your school in two sessions of about an hour, set and scored by the agency.
• The five areas: number; algebra; data; geometry and measurement; financial literacy.
• The four levels: Level 3 is the provincial standard; Level 4 above; Levels 1 and 2 below. Not a percentage, not a rank, not a label.
• It counts: part of the course mark under the Ministry's policy, between 10 and 30 per cent as your school or board sets it. Ask your teacher how, in the first week.
• The only practice test is the agency's own sample test, in the real format.
• The plan: sample test, error log, top three areas, mixed sets and recall for two weeks, the sample again a week before, the last two days light.