Supporting English Language Learners in Statistics and Data Handling (IB MYP)

A student who can calculate a mean, a median and a mode perfectly can still score badly on a statistics assessment in the MYP, because the questions are written in English rather than in symbols. Words like spread, range, represent, consistent and typical carry the entire mark, and they are not taught anywhere. For multilingual learners in Years 2 to 4, this topic is one of the clearest cases where the mathematics is accessible and the language is not.

The words that cause the trouble are not the statistical ones

Mean, median, mode and range get taught explicitly and usually stick. The words that sink students are the ordinary ones being used technically. "Spread" is not about butter. "Represent" does not mean draw. "Consistent" in a comparison question means something very specific about variation. And the comparative constructions the answers require, such as "on average, higher, but less consistent", combine two ideas in a grammatical structure many students have never produced.

Make a list of the twelve non-mathematical words your assessment actually uses. Take them from a past task rather than from a textbook glossary. That list, not the statistical vocabulary, is your language teaching plan for the unit.

Staging vocabulary across the unit

Stage one is receptive. Students meet each word in context, match it to a picture or a dataset, and sort statements as sensible or not sensible. No production required. Two lessons of this is enough.

Stage two is supported production. Students choose from three offered sentences to describe a dataset, then say why the other two are wrong. Choosing before composing is the step teachers most often skip, and it is where the register actually forms.

Stage three is independent production with a frame available but not required. By this point the frame should feel like a safety net rather than a script. Getting there needs volume of practice rather than more explanation, and a graduated bank such as Interactive Math Workshop – Multiply, Divide & Measure with Confidence is useful because the calculation load stays low enough that the student's attention is free for the sentence they are writing.

Scaffolds that survive contact with an assessment

Three that carry their weight. First, a comparison mat: two columns, one per dataset, with fixed prompts for average, spread and one thing that stands out. Students fill the mat before writing prose, which converts a writing task into a transcription task.

Second, a question decoder. Students underline the command term, circle the dataset named, and box any word that tells them which measure is wanted. Three marks, no maths yet. Multilingual learners lose marks far more often to misreading the question than to miscalculating.

Third, worked answers with the language annotated rather than the arithmetic. Show a full-mark response and label why "however" appears where it does. Students are shown correct answers constantly and are rarely shown why they are written that way.

Choosing data that does the language work for you

Context reduces language load when students already know something about the subject, and increases it when they do not. Datasets about the class, the school day or local weather are safest early on. Datasets requiring cultural background knowledge, such as national sports statistics, quietly disadvantage a recently arrived student twice over.

Once the register is secure, a context worth arguing about is what pushes students into the harder constructions, because they have something to say. Climate Change Data Lab – Understanding Climate Change Through Real Data supplies datasets where the comparison genuinely matters, which is when a student stops producing minimal answers and starts reaching for language they did not have last month. It also sits naturally under a global context of globalisation and sustainability, so the statement of inquiry does not have to be invented separately.

Marking so the mathematics is visible

Split your marking into two passes. First pass, the mathematics only: is the measure correct, is the comparison valid, is the conclusion supported. Second pass, the language, in a different colour, and limited to two corrections. A page returned with every article and preposition corrected teaches a student that the subject is English, and they will write less next time to reduce the risk.

Keep an eye on where the topic is going, too. Students who build the comparison register now are the ones who cope when this content becomes formal inference in the final years, where the language of evidence gets more demanding rather than less. A senior treatment such as Hypothesis Testing & Errors – Full Statistics Unit (Grades 11–12) is worth looking through for that reason: it shows you the sentences your MYP students are eventually being prepared to write, which sharpens what you insist on now. The target for this unit is straightforward. A student compares two datasets aloud, uses "however", and is right.

Related reading from the Teacher Hub

• Algebraic Manipulation in Five Lessons: A Sequencing Guide (IB MYP)
• Group Work That Actually Works for Portraiture (IB MYP)
• Marking Art History Faster Without Losing Quality (IB MYP)