Why This Matters

"So a level 6 out of 8 is basically 75%, right?"

You've probably asked yourself some version of this, especially if you learned to grade in a system built on marks out of 100 — including report-card systems like Vietnam's MOET framework, where a score of 8/10 has an unambiguous, portable meaning. In MYP, that instinct will mislead you constantly, and it will mislead you in a way that's hard to notice because the numbers look like percentages. They aren't. Treating them as if they were is the single most common distortion new MYP teachers bring into their marking, and it quietly changes the judgements you make without you realizing it's happening.

Understand the Idea

There are, broadly, three different logics for turning student work into a mark, and it's worth naming all three so you can see what MYP is asking you to do differently.

Norm-referenced grading ranks students against each other — your grade depends partly on how your peers performed. A curve is the clearest example. MYP does not work this way.

Percentage/points-based grading starts from a total (say, 100 points) and subtracts for errors or omissions: lose 2 points for a missing citation, lose 5 for a weak thesis. The final number is a running total of deductions. This is common in many national systems and it produces a very specific mental habit: read for what's wrong, subtract, add up.

Criterion-related assessment — what the MYP uses — asks a different question entirely: which achievement level descriptor, as a whole, best matches the quality of what this student produced? An achievement level is a band of quality (often numbered something like 1–8, though the exact scale varies by criterion and should be checked against your current subject guide) described in prose, not built from a point total. You are not deducting from a perfect score. You are matching the work, as a whole, to the descriptor that fits it best — this is called best-fit judgement: choosing the level whose description most accurately captures the work overall, even if the work doesn't tick every single clause perfectly.

This is a genuinely different cognitive act, not just a relabeling. Points-based grading is subtractive and additive — you start at the top (or zero) and move by increments. Best-fit judgement is comparative and holistic — you hold the work next to several descriptors and ask which one it resembles most.

Why does this matter enough to be its own article? Because if you privately convert in your head — "level 6 of 8 feels like about 75%, so I'll report it like a B" — you import the logic of deduction into a system that was never built to be linear or evenly spaced. A jump from one level to the next does not represent a fixed, equal amount of quality, the way going from 75% to 80% does. Levels are qualitative bands, not points on a ruler. 📌 Official source: how your school converts achievement levels into any reported grade (if it does at all) is governed by current IB guidance and/or your school's own reporting policy — check both rather than assuming a fixed formula exists.

⚠️ Common trap: Building your own informal conversion table ("1–2 = F, 3–4 = D," and so on) and using it consistently enough that it starts to feel official. It isn't, and it will produce judgements that don't match a colleague's — because your colleague is (correctly) not doing this at all.

See It in Practice

Take one piece of student work — a persuasive essay from a Grade 8 Language and Literature class — and watch it get marked two different ways.

Point-deduction approach (the habit many teachers bring in):

"Started at 100. Thesis is present but vague — minus 8. Two paragraphs lack topic sentences — minus 10. Good use of a counter-argument — no deduction, actually add nothing since we're only deducting. Conclusion restates rather than extends — minus 7. Grammar mostly clean, two agreement errors — minus 3. Final score: 72/100 → converted in my head to 'about a level 5.'"

Notice what happened: the teacher generated a number nobody asked for, then reverse-engineered a level from it. The level wasn't chosen by matching the work to a descriptor — it was guessed backward from an invented percentage.

Best-fit judgement approach (what criterion-related assessment actually asks for):

"Reading the relevant level descriptors for this criterion, band 5–6 describes an argument that is generally clear with adequate development and some use of counter-argument; band 7–8 describes an argument that is consistently clear, well developed, and effectively uses counter-argument. This essay has a real but underdeveloped thesis, inconsistent paragraph structure, and one genuinely effective counter-argument. That's a stronger match to the 5–6 band than the 7–8 band — the argument isn't consistently well developed. Best fit: level 6, upper end of that band, because the counter-argument use is a real strength pulling it up within the band."

Same essay. Different process. Different kind of justification — one you could actually defend to the student, in writing, using the descriptor's own language, rather than a hidden arithmetic trail nobody else can check.

🧠 Teacher insight: The tell that you're still deduction-thinking: if you can hear yourself saying "lost marks for," stop and reread the descriptor instead. "Lost marks for" is percentage language wearing a criterion costume.

Apply It

🛑 Do this now — the conversion trap exercise.
  1. Pick one real (or realistic, invented) piece of student work from a class you teach.
  2. Mark it fast, by instinct, as if it were out of 100 — the way you might have been trained to. Write down the number and the two or three things you deducted for. Don't overthink it; this is meant to surface your default habit, not produce a polished score.
  3. Set that number aside. Now open the actual current level descriptors for the relevant criterion (from your subject guide) and mark the same piece of work again, properly — matching the whole piece to the band that fits best, strand by strand if the criterion has strands (see Article 21).
  4. Write two or three sentences comparing the two outcomes. Where did they diverge? What did the percentage instinct cause you to over- or under-weight? Was there a moment where a "deduction" you made in step 2 doesn't actually correspond to anything in the real descriptor language?

Artefact to produce: A short written comparison (your percentage guess vs. your descriptor-matched level, with the gap explained in your own words). Keep this — it's a genuinely useful thing to reread before every major marking session in your first year, because the instinct doesn't disappear after one exercise. It fades with repetition.

Avoid This

Silently building a personal conversion chart. Even if you never write it down, if you consistently think "level 7 = A, level 5 = C" you're back to percentage logic with extra steps. Fix: mark from the descriptor language every time, not from a memorized equivalence.

Assuming a higher level always means "more correct answers." Levels describe quality of thinking and skill, not quantity of correct facts. A shorter response that reasons more clearly can outrank a longer one that lists more information. Fix: reread what the strand's command term is actually asking for (Article 21).

Reporting decisions built on an assumed percentage-to-level table. Different schools and different moments in the IB's own guidance may treat grade determination differently. Fix: don't assume — ask your Head of Department or MYP Coordinator what your school's current reporting policy actually says, and check current IB guidance yourself. 📌 Official source: your school's assessment policy document and current IB guidance on grade/level determination.

Check Yourself

Can you explain, in one or two sentences and without using the word "percent," why a level 6 is not "75% of a level 8"? If your explanation still leans on a hidden conversion, reread "Understand the Idea" above before moving on.

Continue Learning

Next step → Article 23 shows you how to take these same published criteria and make them concrete for one specific task — without changing what they say. Downloadable resource: Percentage vs Criterion mental-model card. 📌 Official source: check your school's current assessment and reporting policy, and current IB guidance on determining achievement levels, before finalizing how you report grades.