The Parent's Playbook

How to support your child through IB Maths — without turning it into a battleground

Practical, honest guidance from an IB Maths examiner and Head of Maths who has watched hundreds of DP students go from panic to composure. Nothing sales-y. Just the eight things I wish every parent knew.

1. Should they take AA or AI? SL or HL?

This is the question I get asked most. The honest answer is: it depends on where your child wants to go to university, not on which one is "harder".

Analysis & Approaches (AA) is the algebraic, proof-heavy course. It's a hard requirement for competitive engineering, physics, pure maths, and increasingly for computer science and economics degrees. If your child is thinking Oxford, Cambridge, Imperial, MIT, or any competitive STEM programme — check the specific course prospectus. Most will list AA HL as required or "strongly preferred".

Applications & Interpretation (AI) is the applied, statistics-and-modelling course. Better for data science, business, psychology, biology, environmental science, and design. It uses the graphical calculator far more heavily than AA does.

SL vs HL is about depth — HL is 240 taught hours vs SL's 150, so significantly more content, plus a compulsory Paper 3 investigation. Pick HL if their target degree needs it; pick SL if their other five subjects already have three HLs. Fewer, stronger grades usually beat a stretched cluster of six.

My rule of thumb: if you're still unsure after checking the university prospectuses, go and sit with the maths teacher for 20 minutes. They already know your child's mathematical fluency better than a course-comparison chart ever will.

2. The university reality check — where SL AI is (and isn't) accepted

This is the section I wish every parent read before their child chose their maths course, not after. Applications & Interpretation SL is a good course for the right student — but for a large slice of degree programmes worldwide, it will quietly close doors that the family didn't realise were closing. I've watched too many DP2 students in April panic-emailing admissions offices because the required maths on the offer letter didn't match what they were sitting.

Here is the honest picture, region by region, current for the 2026 admissions cycle.

United Kingdom — the strictest market for SL AI

UK universities are the most explicit in the world about what they will and won't accept. Imperial College London publicly states a strong preference for Analysis & Approaches across all STEM courses, and lists AA HL as the specific requirement for a number of engineering and computing programmes. Oxford, Cambridge, LSE, UCL, Warwick and Bristol typically require AA HL (grade 6 or 7) for engineering, physics, mathematics, economics (at the top schools) and computer science. Manchester and Edinburgh occasionally accept AI HL for the same subjects but the default is still AA HL.

Specifically, Cambridge Engineering typically requires a 7 in AA HL. Imperial Engineering asks for a 6 or 7 in AA HL. LSE Economics wants a 7 in AA HL. AI SL is not accepted for any of these specific courses. AI HL is accepted at LSE for a handful of non-quantitative social science programmes (check the individual course page for the year of entry), but for the flagship quantitative degrees it isn't.

Where SL AI is fine in the UK: most humanities, most social sciences, law, English, history, modern languages, media, business at non-Russell-Group universities, art and design, and vocational courses. If your child's target list is those subjects, SL AI is a perfectly sensible choice.

European technical universities — often a hard "no"

Continental Europe's engineering-first universities are frequently even stricter than the UK. Here is the pattern you actually encounter on offer letters:

• TU Delft (Netherlands) — explicitly refuses Applications & Interpretations at any level for its BSc engineering programmes. AA HL + Physics HL is required. Aerospace, Mechanical, Civil, Computer Science & Engineering — all closed to AI students, even AI HL.
• Eindhoven University of Technology (TU/e) and University of Twente have similarly restrictive policies — HL maths (typically AA) is required for engineering programmes, and AI SL is not sufficient.
• ETH Zurich (Switzerland) — will accept either AI or AA but only at HL. AI SL is not sufficient for direct entry, though AA HL is by far the strongest signal for competitive programmes like Computer Science & Engineering.
• EPFL (Lausanne) — similar policy; HL maths required, AA strongly preferred.
• TU Munich, RWTH Aachen, KIT (Germany) — competitive engineering programmes typically expect HL maths, with AA HL the preferred combination. Access rules vary by state so check the specific programme page.
• Trinity College Dublin & UCD (Ireland) — engineering entry typically expects AA HL; AI is generally not preferred for these programmes.
• KTH (Sweden), DTU (Denmark), NTNU (Norway) — HL maths required for engineering; AA preferred.

Where SL AI is accepted across Europe: business schools (Bocconi, ESADE, HEC, IE), most humanities and social science courses at general universities, arts, design, and international relations. If your child is aiming at a Dutch University College or a broad liberal-arts programme (Utrecht, Amsterdam UvA, Maastricht), SL AI is fine.

United States — technically accepted, competitively fatal for STEM

US universities are officially "holistic" — they don't publish hard maths requirements the way the UK does. What that means in practice is subtler and, honestly, more dangerous: SL AI is not banned, but for competitive STEM it is a serious handicap.

MIT, Stanford, Caltech, CMU (SCS), Princeton, Harvard, and the Ivies for engineering/CS all expect to see the most rigorous maths available at your child's school. If your school offers AA HL and your child takes AI SL, admissions will read that as a signal — they didn't stretch. Compensating requires exceptional external evidence: SAT Maths 780+, AMC/AIME participation, math olympiad medals, university-level maths courses. Most students can't muster that.

Where SL AI is fine in the US: everywhere for non-STEM majors — humanities, social sciences, business (except Wharton and MIT Sloan), pre-med at most colleges (though top med schools then want strong SAT Maths and Calc), design, arts, communications. Most liberal-arts colleges (Amherst, Williams, Bowdoin) will accept SL AI without prejudice for any non-STEM major.

Canada — increasingly strict on STEM

Waterloo (Engineering, CS, Maths) typically expects AA HL and is not a realistic target with AI SL. Toronto's Engineering Science and Rotman Commerce list AA HL as the expected qualification. McGill Engineering typically expects AA HL. UBC Engineering accepts AA HL or SL, but AI SL is generally not sufficient. For non-STEM programmes at Canadian universities, SL AI is fine.

Australia & New Zealand — mostly accepting, with STEM caveats

Australian universities are more flexible than the UK or Netherlands. Melbourne, UNSW, Sydney, Monash, ANU — all publish IB entry requirements that usually accept AI SL for general degrees. But for Engineering, Computer Science, or the Bachelor of Science with a physics/maths major, they'll typically require HL maths (either flavour), and Melbourne's premium programmes (Chancellor's Scholars, Engineering Honours) will effectively expect AA HL. Auckland and Otago in New Zealand are similar.

Asia — heavily filtered on maths course

• NUS & NTU (Singapore) — competitive engineering, computing, and business analytics programmes strongly prefer AA HL. AI SL applicants often need bridging courses or are quietly filtered out even if the published minimum is met.
• HKUST, HKU, CUHK (Hong Kong) — engineering and business (BBA) programmes want HL maths; AA preferred.
• IITs, IISc, BITS (India) — most IITs admit via JEE only, so IB entry to Indian engineering is limited. IB-friendly options like Ashoka, Krea, and O.P. Jindal accept SL AI for non-STEM.
• Japan (Todai, Kyoto U, Waseda SILS, Keio) — HL maths expected for STEM; AI SL usable for social sciences and international programmes.
• South Korea (SNU, KAIST, Yonsei UIC) — HL maths expected for STEM.

Middle East

NYU Abu Dhabi and NYU Shanghai use the same standards as NYU New York — SL AI is fine for non-STEM, weaker for STEM. Khalifa University (UAE) and American University of Sharjah expect HL maths for engineering (AA preferred). American University of Beirut accepts SL AI for most non-engineering degrees. Note: KAUST is a graduate-only research university and doesn't take undergraduates, so it isn't a factor here.

Quick decision table — what your child needs to take

Rule of thumb, current for the 2026 cycle. Individual courses vary — always check the specific degree page on the university's own website, not a third-party summary.

Switching courses mid-DP: if you're reading this in DP1 and realising your child is on the wrong course, most schools will let a student move down (HL → SL, AA → AI) up until January of DP1. Moving up is much harder and usually needs a summer of self-study plus a re-assessment. Moving from AI SL to AA SL mid-DP1 is generally possible if the school is on-side; moving from AI SL to AA HL is a huge stretch and rarely worth attempting after October.

3. Maths anxiety is real — and it's contagious

The single biggest predictor of a student's maths confidence isn't natural ability. It's what they hear at home. If you ever say "I was never good at maths" or "your dad's better at maths than me", your child hears "maths is a fixed genetic trait" — and if they've already decided they're on the wrong side of that trait, they'll stop trying.

4. Praise the effort, not the grade

When a child gets a 7 on a test and you say "amazing, you're so smart", you've quietly told them that intelligence is fixed and grades measure it. When they get a 4 next time — because DP is harder than pre-DP — they'll conclude they've become less smart.

Praise the specific thing they did: "I noticed you re-drew that diagram three times until you were happy with it. That's what strong maths students do." That's a growth-mindset praise. It's the difference between a child who bounces back from a bad mock and a child who gives up.

5. The IA is 20% of the final grade — and it's the most under-prepared part

The Internal Assessment (the "IA" or "exploration") is a 12–20 page maths investigation your child chooses. It counts for 20% of the final grade across all four maths courses. Most students leave it far too late.

What actually helps: get them to try our free IA Topic Explorer in the first term of DP2 — that gives them a shortlist of realistic topics before they burn a week on something unworkable. Then help them stick to a schedule of one page a fortnight from October.

Whatever you do, do not write it for them or edit the mathematics. Examiners can see it a mile off. What you can do is proof-read for typos, ask them "why did you choose this method?", and be a rubber duck.

6. Tutoring is expensive. Structure isn't.

A private IB Maths tutor typically costs €40–80 per hour, so 90 minutes a week for two years quietly totals €4,000+. Before you commit to that, ask what's actually missing: is it knowledge (they haven't been taught something), or practice (they've been taught but can't recall under pressure), or confidence (they know it but freeze)?

Only the first genuinely needs a tutor. Practice and confidence are what a good revision engine — worked examples, AI feedback on their own scribbles, and a clear "path to 7" — actually solves for a tenth of the cost. If a tutor IS the right answer, treat it as a short intervention, not a two-year subscription.

7. Predicted grades matter more than mock grades

Universities offer places on the predicted grade, not the mock. Predicted grades come from the school and are based on classwork, mocks, IA progress, and the maths teacher's professional judgement. If the predicted grade lands too low for your child's target course, that's the fire drill — not the mock result.

How to lift a predicted grade: consistent IA drafts submitted on time, strong participation, and a visible improvement trajectory. Teachers weight all three heavily. A student who has been at 5 for two years but arrives at Christmas with a 6 in the mock will often get predicted a 6, because the direction of travel is right.

8. Screens, sleep, and the last 6 weeks

DP1 is a marathon. Sleep matters more than any tutoring session. Consolidation of maths procedures happens overnight — a student sleeping five hours cannot retain what they revised. Set a phone curfew if you can (many families settle on 10pm phone in the kitchen). Sleep will move their grade more than anything I could ever teach them.

In the final six weeks before May exams, the pattern is: past papers, mark them, re-do the ones they got wrong. It's not glamorous — but that specific loop consistently lifts a student one grade band. Not two hours of revision reading; two hours of past-paper writing followed by 30 minutes of marking.

9. When to worry — and when not to

A student getting Cs in mid-DP1 is normal — pre-IB Maths is a big step up and it takes a full term to recalibrate. A student getting Es in their second mock exam or refusing to attempt papers is a different signal. That's the point to sit with their maths teacher, and if the response is "I'm doing everything I can", consider whether it's anxiety, an underlying learning difference (dyscalculia is real, though rare — around 5% of students), or a course-choice mismatch (an AA HL student who should have taken AI SL is not unusual).

Almost every failing DP maths student I've taught was struggling with anxiety, not ability. Treat the anxiety, and the maths tends to follow.