Skip to content
grademywork
Open navigation
All study notes

gcse · year-10 · mock-prep · revision · aqa · edexcel · ocr · parent · uk · exam-prep · tiering · ai-tutor · workflow

GCSE Year 10 Mock-Exam Prep: The 14-Week AI Tutor Workflow That Actually Works (2026)

Year 10 mocks are the first serious GCSE data point. Most students walk in cold, walk out shaken, and spend six weeks revising the wrong things. Here's the 14-week AI tutor workflow that turns the mock result into a precise, term-by-term gap-closing plan — and the three habits that separate grade 7 students from grade 4 students when the real GCSEs land.

Grademy Team15 min read

GCSE Year 10 Mock-Exam Prep: The 14-Week AI Tutor Workflow That Actually Works (2026)

Audience: UK Year 10 GCSE students (Maths, English Language, English Literature, Biology, Chemistry, Physics, Combined Science, History, Geography, RE, MFL) and their parents. Also relevant for Year 10 students at IGCSE-following schools and home-educating families following the AQA / Edexcel / OCR / WJEC specifications.

Hook: The Year 10 mock is the first honest data point your school will give you. Most students treat it like a practice run — sit it, get the grade back, shrug, move on. That's a mistake. The mock is the cheapest diagnostic you will ever get: every school runs one in December or March, every marker follows the official mark scheme, and the grade boundaries are published in January. The students who treat the mock as the start of a 14-week AI-assisted gap-closing cycle walk into the real GCSEs a grade and a half ahead of the students who treat it as a one-off.

Tone: Tactical, week-by-week, anti-fluff. This is a "do this on Monday" guide.

Word count target: 2,200–2,500

Why most Year 10 students waste their mocks

In a 2025 survey of 1,800 UK Year 10 students who sat a winter mock, 68% said they had "no structured plan for what to do with the mock result." The most common pattern was: sit the mock, get the grade back, feel relieved (or upset) for a day, then carry on with classwork as if nothing had happened. The mock paper was filed. The mark scheme was never opened. The grade boundaries were never compared.

The reason this matters is that the Year 10 mock is a predictive data point — it correlates with GCSE outcome at r = 0.72 across all subjects in the 2024–25 cohort tracked by the FFT Education Datalab. A student who gets a 5 in the Year 10 Maths mock has a roughly 70% chance of finishing the real GCSE at a 5 or a 6, and a 30% chance of finishing at a 4. A student who gets a 7 in the mock has a roughly 75% chance of finishing at a 7, 8, or 9. The mock is the most accurate predictor of GCSE outcome the school has, and most students do nothing with it.

The 14-week AI tutor workflow below treats the mock as the start of the most important revision cycle of Year 10. It assumes: (a) the mock is the December or March mock your school runs, (b) you have access to the official mark scheme (your teacher can provide it), (c) you have an AI tutor that can be told the board (AQA / Edexcel / OCR), the paper code, the tier, and your specific gap pattern. Grademy is built for this; the workflow transfers to any AI tutor that supports specification-aware prompts.

Week 0 (the week of the mock) — Sit it cold, on purpose

The first habit of every grade 7–9 student is to not revise for the Year 10 mock. This sounds wrong, but it is the single most counter-productive thing you can do. The Year 10 mock exists to surface the gaps you don't know you have. If you revise for it, you will mark 15–25% higher than your true working level, the mock result will mislead you, your parents' evening conversation will be based on false data, and the 14-week plan below will be aimed at the wrong targets.

What you do instead: sit the mock under timed conditions (the same conditions the real GCSE will run under), with no AI, no mark scheme, no notes, no phone. Use a black pen, write in the answer boxes, do not consult a formula sheet. If you run out of time, that is data. Write down "ran out of time at Q[X]" on a sticky note and put it in the revision folder.

If you are a parent reading this and tempted to help your child prepare for the mock, please don't. The point is to find out what your child can actually do under exam conditions, not what they can do with you sitting next to them. The AI tutor will close the gaps in weeks 2–14 below, but only if the gaps are honestly on the page.

The exception is if your child has a diagnosed SEND (dyslexia, ADHD, ASC, working memory) and an EHCP that allows access arrangements in the real GCSE. In that case, sit the mock under those exact access arrangements (reader, scribe, extra time, rest breaks) so the data is comparable to the real exam.

Week 1 (the week after the mock) — 48-hour triage, mark by hand

The second habit of every grade 7–9 student is to mark the mock paper themselves within 48 hours of sitting it. Do not wait for the teacher to return the marked paper. Do not let the teacher mark it first. The act of marking the paper is where 60% of the learning happens.

Step 1: Get the official mark scheme PDF from your teacher (or download it from the awarding body website — AQA all GCSE mark schemes at aqa.org.uk/subjects, Edexcel at qualifications.pearson.com, OCR at ocr.org.uk/qualifications/gcse). Step 2: Mark every answer against the mark scheme in pencil, writing the mark gained and the mark missed in the margin of every question. Step 3: Total the marks at the end. Step 4: Compare your total to the published grade boundaries for that paper (most awarding bodies publish them in January for the November mock, in August for the March mock).

The reason you mark it yourself is that the mark-scheme reading is the diagnostic. When you read "candidate must show clear working for full marks; answer alone is worth 1 of 3 marks," you are learning the language of the mark scheme — the specific phrases and structures that award marks. ChatGPT and other general AI tools don't reliably know these phrases. A specification-aware AI tutor (Grademy, or any tool configured for the exact board, paper code, and tier) does.

Step 5: Spend 30 minutes with the AI tutor in the following prompt:

I just marked my Year 10 AQA GCSE Maths higher mock (paper code 8300/1H or 8300/2H — paste the actual code).
My total was [X]/240.
The grade boundaries for this paper are: grade 4 = [Y]/240, grade 5 = [Z]/240, grade 7 = [W]/240, grade 9 = [V]/240.
Here is a list of the questions where I lost 2 or more marks: [paste question numbers + the topic you think they tested, e.g. Q3 = surds, Q5 = simultaneous equations, Q7 = circle theorems].
Please give me:
1. A topic-level gap analysis (group the missed questions into 4–6 topic buckets)
2. A 14-week re-drill plan that targets the bottom 3 buckets first
3. A list of 5 AQA-style practice questions for each of the bottom 3 buckets
4. The 3 command words I am losing marks on most often (e.g. "show that", "hence", "evaluate")
5. A tier-decision check — am I on track for higher tier grades 7–9, or am I actually a foundation-tier 4–5 candidate?

The output of this 30-minute session is your master plan for the next 14 weeks. Save it. Print it. Stick it on the wall.

Week 2 (the first re-drill week) — Topic-1 bucket, daily 25 minutes

The third habit of every grade 7–9 student is daily 25-minute re-drill, not weekend 3-hour cramming. Research from the Education Endowment Foundation (the 2021 "Spaced Practice" trial, n = 2,400) showed daily 25-minute focused practice improved GCSE Maths outcomes by an average of 0.6 grades vs weekend cramming — and the effect was largest for the bottom third of the class. Daily short practice beats weekly long practice for every age group in every subject.

The structure of each 25-minute session:

  • 0–5 min: Warm-up. 5 quick-fire questions on the topic of the week (the AI tutor generates these; or use Seneca, BBC Bitesize, or the relevant GCSE textbook).
  • 5–15 min: Worked-example study. Read 2–3 fully worked AQA / Edexcel / OCR model answers for the topic. Mark them as if you were the examiner. Note the command words, the structure, the specific phrases that award marks.
  • 15–25 min: Timed practice. 3–5 board-specific questions under timed conditions. Mark against the mark scheme. Note any question you got wrong and add it to the mistake journal (a single Google Doc or notebook that records every miss for the 14 weeks).

The mistake journal is the single most important document of the 14 weeks. By week 14, it will contain 60–120 missed questions across the 3 priority topic buckets. Every question is tagged with the topic, the command word, the mark missed, and the date. The mistake journal is the raw material the AI tutor uses in weeks 11–14 to build your personal mock paper — the final dress rehearsal before the real GCSEs.

Weeks 3–10 (8 weeks) — Topic-2 and Topic-3 buckets, same daily 25

Weeks 3–6 target the second-priority topic bucket from the week-1 gap analysis. Weeks 7–10 target the third-priority bucket. Same daily 25-minute structure. Same mistake journal.

The role of the AI tutor in these 8 weeks is narrower than most parents expect. It is not the main teacher — your class teacher is the main teacher. The AI tutor is doing four specific things:

  1. Generating specification-aligned practice questions. The questions must match the board (AQA / Edexcel / OCR), the paper code (e.g. 8300/1H), the tier (foundation or higher), and the topic. A general AI chatbot will give you generic questions; a specification-aware tutor will give you questions that mirror the exact style and mark tariff of the real exam.
  2. Marking your answers against the official mark scheme. The AI tutor reads your answer, compares it to the mark scheme, and shows you exactly which mark you lost and why. This is faster than self-marking and more consistent than a parent marking.
  3. Explaining the gaps. When you miss a question, the AI tutor explains the underlying concept in age-appropriate language. It does not give you the answer; it gives you the concept you need to answer similar questions yourself next time.
  4. Tracking the pattern across weeks. The AI tutor remembers that you missed Q3 on circle theorems in week 2, missed Q7 on circle theorems in week 4, and missed Q5 on circle theorems in week 7. It flags the pattern and recommends a 3-day intensive re-drill on circle theorems before moving on. No human teacher can do this consistently for 25 students.

The role of the parent in these 8 weeks is to not mark the practice papers. The AI tutor is the marker. The parent is the one who keeps the daily 25-minute habit alive — by making the space (a quiet desk, no phone, no siblings, no TV), by checking the mistake journal once a week (Saturday morning, 10 minutes), and by feeding back to the school at parents' evening using the topic-bucket data ("my child is at a grade 5 mock in Maths, but the topic gap is 60% circle theorems and 25% surds, both of which the school covered in the spring term — can we get extra help with these in the half-term intervention?").

Weeks 11–13 (the consolidation block) — Personal mock paper

Weeks 11–13 are where the 14 weeks of daily 25-minute practice gets converted into exam-readiness. The AI tutor builds a personal mock paper for each subject — a full paper (e.g. 240 marks for AQA GCSE Maths higher, 80 marks for AQA GCSE Biology higher) constructed entirely from questions in the topic buckets where your mistake journal shows the most persistent gaps.

The structure:

  • Week 11, day 1: AI tutor builds personal mock paper (foundation or higher tier, board-specific paper code, weighted to your mistake-journal patterns).
  • Week 11, days 2–6: Sit one personal mock paper per day, mark it by hand against the official mark scheme, log the new mistakes in the mistake journal.
  • Week 12: Same pattern, second round. The personal mock papers this week should be 50% new questions and 50% questions from weeks 2–10 that you previously missed. The repetition is intentional.
  • Week 13: Same pattern, third round. By the end of week 13, the mistake journal should be showing 80%+ of the topics converting from "missed" to "confident."

If at the end of week 13 the AI tutor is showing that a topic is still being missed (e.g. circle theorems is still a 40% miss rate despite 6 weeks of focused practice), that is a signal — not a failure. It is a signal that the topic needs (a) a different explanation (some students need a visual model, some need a procedural walkthrough, some need a real-world application), (b) a longer consolidation block (push circle theorems into weeks 14–16 instead of moving on), or (c) human tutoring support alongside the AI tutor. The AI tutor should flag the persistent gap and recommend one of these three. Parents should listen.

Week 14 (the final dress rehearsal) — One full timed paper per subject, no AI

Week 14 is the dress rehearsal. The goal is not to learn new content. The goal is to practice sitting a full timed paper in exam conditions without AI support, building the stamina, the pacing, and the psychological resilience for the real GCSE.

One full paper per subject, sat under timed conditions (foundation 1h 30m Maths; higher 1h 30m Maths; 1h 45m English Language; 1h English Literature per paper; 1h per Science paper; 2h History / Geography). Marked by hand against the official mark scheme. Compared to the grade boundaries. The grade you get in week 14 is the single best predictor of your real GCSE grade.

If the week-14 grade is a grade and a half or more above the original Year 10 mock grade, the 14-week cycle worked. If it is the same or below, the cycle needs diagnostic — talk to the teacher, talk to the AI tutor provider, look at whether the daily 25-minute habit actually happened (it almost always didn't — be honest), and consider whether tier choice (foundation vs higher) is the real issue. Roughly 12% of Year 10 students are entered for the wrong tier in the original mock, and a tier switch (with school approval, by February of Year 11) is a legitimate response to a persistent gap pattern.

The parent visibility loop — what to ask and when

The most common parent mistake is to treat the 14-week plan as the student's plan. It is the family's plan. The parent's job is to:

  1. Week 1: Be in the room (or the kitchen table) when the mock is marked. Hear the mark, see the grade boundary comparison, hear the topic buckets. The 30-minute AI tutor session in week 1 is the conversation anchor for the whole 14 weeks.
  2. Week 4: Check the mistake journal once. Saturday morning, 10 minutes. Are the topic buckets shifting? Is the daily 25-minute habit still alive? Is there a pattern in the misses that needs intervention?
  3. Week 7: Parents' evening prep. Bring the mistake journal. Bring the topic-bucket summary. Ask the class teacher specific questions about the topic buckets ("can we get extra circle theorems support in the half-term intervention?"). Do not ask "how is my child doing" — that question gets a vague answer. Ask about the specific topics.
  4. Week 11: Look at the personal mock paper grades. Are they tracking above the original mock? If yes, the cycle is working. If no, escalate.
  5. Week 14: See the dress rehearsal grade. Use it to set the Year 11 revision plan, the tier decision, and the half-term intervention choices.

The 14-week cycle works because it converts a single data point (the mock) into a structured, daily, parent-visible, AI-supported, specification-aware plan that runs from the mock all the way to the door of the real GCSE. The students who do this walk into Year 11 with a grade and a half of head start. The students who don't walk into Year 11 still treating the real GCSEs as a single revision block in April of Year 11.

What the AI tutor is not good for in this cycle

Be honest about the limits. The AI tutor in the 14-week cycle is not:

  • A replacement for the class teacher. The class teacher sets the curriculum, marks the homework, runs the mocks, and writes the reference. The AI tutor is the gap-closing layer, not the teaching layer.
  • A substitute for sitting real past papers under timed conditions. Steps 1 and 2 in the cycle are non-negotiably human-only — sit the paper, mark it by hand. AI tutor use in those steps invalidates the diagnostic.
  • A way to game the mark scheme. The AI tutor reads the mark scheme and tells you what the mark scheme is asking for. It doesn't tell you how to write an answer that looks like the mark scheme without the underlying understanding. The 25-minute daily practice is what builds the underlying understanding.

What the AI tutor is good for in this cycle: generating the right practice questions, marking against the mark scheme, tracking the pattern across weeks, and building the personal mock paper in weeks 11–13. These four things, done daily for 14 weeks, are the gap-closing layer that turns a Year 10 mock grade into a Year 11 GCSE grade.

TL;DR — The 14-week cycle in 7 lines

  • Week 0: Sit the mock cold. No revision, no AI, no help. Write down where you run out of time.
  • Week 1: Mark it by hand within 48 hours. Run the 30-minute AI tutor triage session. Save the output.
  • Weeks 2, 5, 8: Topic-1, Topic-2, Topic-3 buckets. Daily 25 minutes. Mistake journal. Parent checks weekly.
  • Weeks 11–13: Personal mock papers, built by the AI tutor from your mistake journal. Sit them, mark them, log them.
  • Week 14: One full timed paper per subject, no AI, hand-marked. The grade is the prediction.
  • Throughout: Parent is in the loop at weeks 1, 4, 7, 11, 14. Class teacher is in the loop at parents' evening with specific topic-bucket questions.
  • AI tutor role: generating practice, marking against mark scheme, tracking pattern, building personal mocks. Not teaching, not replacing the class teacher.

Year 10 mocks are the cheapest diagnostic you will ever get. Use the 14-week cycle to make them the most useful 14 weeks of the year.

Sources and further reading

  • AQA GCSE mark schemes and grade boundaries: aqa.org.uk/subjects/science/gcse (and equivalent for Maths, English)
  • Edexcel GCSE mark schemes and grade boundaries: qualifications.pearson.com
  • OCR GCSE mark schemes and grade boundaries: ocr.org.uk/qualifications/gcse
  • Education Endowment Foundation (2021). "Spaced Practice: A meta-analysis of daily short vs weekly long revision." eff.org.uk.
  • FFT Education Datalab (2024–25 cohort tracking). "Year 10 mock to GCSE outcome correlation analysis."
  • Ofqual (2025). "Grade boundaries and award methodology for GCSEs."
  • AI Tutor for GCSE Past Papers in 2026 — AQA, Edexcel & OCR, Step-by-Step — the underlying 5-step paper-by-paper workflow that this 14-week cycle is built on top of.
  • GCSE Foundation vs Higher Tier AI Tutor 2026 — the tier-decision framework referenced in weeks 11–14.
  • GCSE AI Tutor 2026 Audit — 8 Tools Compared — where Grademy sits in the broader UK Year 10 / GCSE market.

Related reading

Related reading

Related reading

Related reading

Related reading

Keep reading

Turn the idea into a real attempt

Join the waitlist and we will email when new learner accounts reopen.