AQA GCSE Statistics (8382) Higher tier is the entry that awards grades 4 to 9, with an allowed grade 3 for a student who just misses grade 4; both 80-mark papers are sat at one tier in one series. By our count from the specification layout, 30 of its 72 content references carry Higher-column text and 11 carry text only there. AQA lists Higher Paper 1 for 7 June 2027.
This page sets out the Higher tier from AQA’s own documents: the specification, its grade boundary documents, and the June 2025 reports on the examination for Higher Paper 1 and Paper 2. The whole qualification, including its assessment objectives, is on our AQA GCSE Statistics page, and the two-year plan is on our Year 11 AQA GCSE Statistics page.
Book a free consultation to talk through Higher tier AQA GCSE Statistics tuition for a Year 11 student.
AQA’s specification states the tier structure in one sentence: “GCSE Statistics has a Foundation tier (grades 1 – 5) and a Higher tier (grades 4 – 9).” Of the Higher tier it then says: “A student taking Higher tier assessments will be awarded a grade within the range of 4 to 9.” The grade below that has its own rule: “A student sitting the Higher tier who just fails to achieve grade 4 will be awarded an allowed grade 3.” Under the minimum for that allowed grade, the specification says the student is recorded as U (unclassified) and receives no qualification certificate.
Entry is for one tier. The specification says: “Students can only be entered for foundation or higher tier not both.” It adds: “Students must take two question papers at the same tier. All question papers must be taken in the same series.”
The two papers are each a written exam of 1 hour 45 minutes, 80 marks, and 50% of the GCSE. Each carries multiple choice and short answer questions plus a Statistical Enquiry Cycle (SEC) question, and the specification lists the content assessed on each as all specification content. The marks add to a total scaled mark of 160, and the specification says “Grade boundaries will be set using this total scaled mark.” Every boundary in the table below is therefore out of 160.
The content rule is the reason Higher is more than a grade range. AQA writes: “All content can be assessed on Higher tier question papers.” For the other tier it writes: “The content in the 'basic foundation content' column and 'additional foundation content' column can be assessed on Foundation tier question papers.” The specification quotes the Department for Education expectation for the third column: “only the more highly attaining students will be assessed on the content identified in the 'higher content' column.” The grade 1 to 5 entry has its own page: Foundation tier AQA GCSE Statistics.
Which content is only on the Higher tier?
Every content reference in sections A to E of the specification is laid out in three columns: basic foundation content, additional foundation content and higher content. We counted the references by where their text sits on the page. These are our counts from the specification layout, not AQA figures: 72 references (3 in section A, 17 in B, 7 in C, 9 in D and 36 in E); 30 of them carry text in the Higher column (none in A, 3 in B, 3 in C, 7 in D and 17 in E); and 11 carry text in the Higher column and nowhere else.
Reference
What the Higher column says
First row naming it in AQA’s scheme of work
D6
“Apply formula to determine Spearman’s rank correlation coefficient.”
Year two, Term one, week 7
E9b
“Interpret given Pearson’s product moment correlation coefficient in the context of the problem.”
Year two, Term one, week 7
E9c
“Understand the distinction between Spearman’s rank correlation and Pearson’s product moment correlation coefficients.”
Year two, Term one, week 7
E10a
“Comment on the differences between experimental and theoretical values in terms of possible bias.”
Year two, Term two, week 4
E10b
“Know and interpret the characteristics of a binomial distribution.”
Year two, Term two, week 5
E11a
“Know and interpret the characteristics of a Normal distribution.”
Year two, Term two, week 6
E11b
For a Normal distribution, knowing that “values more than three standard deviations from the mean are very unusual”.
Year two, Term two, week 6
E11c
“Use action and warning lines in quality assurance sampling applications.”
Year two, Term two, week 8
E11d
“Use calculated or given means and standard deviation to standardise and interpret data collected in two comparable samples.”
Year two, Term two, week 7
E12d
“Apply Petersen capture/recapture formula to calculate an estimate of the size of a population.”
Year two, Term two, week 11
E13b
“Know that a set of sample means is more closely distributed than individual values from the same population.”
Year two, Term two, week 12
The 11 are listed in the specification’s own words. AQA publishes a two-year scheme of work among its teaching resources, and describes it as “designed as a suggestion only and not as a prescriptive approach”. It assumes “a 60 week course over two years”. Read as a suggestion, it places all 11 in Year two: 3 in Term one (D6, E9b, E9c) and 8 in Term two. Across the 30 references with Higher-column text, our reading of the same document puts the first row for 11 in Year one and for 19 in Year two.
The other 19: Higher text that extends a foundation reference
Nineteen further references have Higher-column text sitting beside foundation-column text. In data collection and design these are multivariate data (B1a), minimising data distortion including level of control (B2d) and the use of control groups (B5c). In charts they are comparative pie charts and comparative 2D and 3D representations (C1b), frequency density for histograms of unequal width (C2) and recognising where misusing the frequency density formula misrepresents a histogram (C3b).
In averages and spread they are the weighted mean, geometric mean and mean seasonal variation (D1a), skewness by calculation (D2), the interpercentile range, interdecile range and standard deviation (D3a), identifying outliers by calculation (D3b), moving averages (D4) and using the equation of the regression line (D5). In interpretation they are comparisons using standard deviation (E3b and E3c), interpreting skewness identified from calculation (E5a), using trends to make predictions (E6), weighted index numbers (E7c), the idea that multiple factors may interact (E8c) and interpreting a calculated Spearman coefficient (E9a). D3a is in this group rather than the 11 because the Basic column carries text for it too (range and quartiles).
Book a free consultation to talk through the Higher tier AQA GCSE Statistics content with a tutor.
What mark did Higher need for grade 4 in each series?
AQA publishes one boundary per grade for each tier, out of the 160 total scaled mark, in its grade boundaries page and grade boundaries archive. The table gives the Higher row (8382H) from the document for each of seven series, every grade column as printed: grades 9, 8, 7, 6, 5, 4 and 3. The last two columns are our calculation, with the working below the table.
Series
9
8
7
6
5
4
3
Our calculation: grade 4 as a share of 160
Our calculation: marks from grade 3 to grade 4
June 2019
139
126
114
94
74
55
45
34.4%
10
November 2021
129
114
99
80
61
42
32
26.3%
10
June 2022
113
97
82
66
51
36
28
22.5%
8
June 2023
128
111
95
76
58
40
31
25.0%
9
June 2024
135
116
98
79
60
42
33
26.3%
9
June 2025
129
112
95
76
57
38
28
23.8%
10
June 2026
114
98
83
67
51
35
27
21.9%
8
Working, taking June 2026: grade 4 is 35 marks, and 35 ÷ 160 = 21.9%. The band between the grade 3 boundary (27) and the grade 4 boundary (35) is 35 − 27 = 8 marks. For June 2019 the same steps give 55 ÷ 160 = 34.4% and 55 − 45 = 10 marks. Between those two series the grade 4 boundary is 20 marks lower (55 − 35) and the grade 9 boundary is 25 marks lower (139 − 114), by our calculation.
The table gives subject boundaries only. AQA presents its notional component grade boundaries “for illustrative purposes only”, so no paper-level boundary is quoted here. Each figure belongs to its own series; none is a prediction for another.
Book a free consultation to talk through Higher tier AQA GCSE Statistics tuition for a Year 11 student.
My son was low in confidence after COVID and the gaps in learning were building up. Using this service helped plug those gaps, rebuild confidence, and improve his ability to learn properly again.
NI
Nicola
★★★★★
Amazing tutors!
We arranged maths and English tuition ahead of GCSEs and both tutors were brilliant. My daughter's grades improved, and by results day she achieved two grades higher than in her mock exams.
S
Saifan Khan
★★★★★
Great Teaching and Focused Learning
I attended an English bootcamp with Leading Tuition and the experience was excellent. The teacher gave focused attention to the small group while keeping the lessons interactive and engaging.
Where did Higher candidates lose marks in the latest reports?
AQA’s assessment resources page lists the June 2025 reports on the examination for Higher Paper 1 and Higher Paper 2, published on 10 July 2026. We read both end to end, part by part: 41 parts in the Paper 1 report and 44 in the Paper 2 report, 85 in all. The marks for each part come from the June 2025 question papers, and each paper’s parts add to 80.
Our counting rule
A part is counted when the report’s own sentence about it uses a negative verdict (not well answered, poorly answered, less successful, challenging) or, where it gives no verdict word, says that fewer than half succeeded or that few, very few or a small share did. A part the report calls well, fairly well, reasonably well or quite well is not counted, whatever mistakes the report lists inside it. By this rule 14 of the 41 Paper 1 parts and 22 of the 44 Paper 2 parts are counted: 36 of 85. They carry 28 marks on Paper 1 and 35 on Paper 2, which is 63 of 160 marks, or 39.4% (our calculation: 63 ÷ 160).
Some parts were read and not counted. Paper 2 Q7(b)(i) (exactly half succeeded), Q8(a) and Q10(a) (both called reasonably well answered), Paper 2 Q13(c)(ii) (the report’s “Again, more than half of the students were unsuccessful” sits between its entries for (c)(ii) and (c)(iii) and is not tied to one part), Paper 1 Q6(a) (called well answered although a significant number gave wrong quartiles) and Paper 1 Q8 (no verdict) are examples. This is one series of reports, so it is a reading of June 2025 and not a forecast.
The result by specification column
For each counted part we asked which specification column holds the skill the examiners describe, using the question paper to see what was asked. This is our classification. Where two readings fall in different columns the part is left unassigned: there are 5 such parts, shown in the table below.
Specification column
Counted parts
Marks they carry
Basic foundation content
7
10
Additional foundation content
9
15
Higher content only
15
28
Not assigned (5 parts, listed below)
5
10
All counted parts
36
63
Our reading: 16 of the 31 assigned parts (7 + 9) sit in foundation-column skills and 15 in Higher-column skills, so about half of the assigned parts are skills a Foundation paper can also assess, and they carry 25 marks (10 + 15) against 28 for the Higher column. The specification says all content can be assessed on Higher papers, so both groups can appear.
What the examiners said about each paper
The Paper 1 report says: “Questions which were not answered well included: drawing and interpreting a cumulative frequency step polygon, identifying outliers by calculation, and determining the reliability of a normal distribution when given the skew and calculating basic probabilities.” It also lists what went well: “Questions that were answered well included: drawing a box plot, calculating a 4-point moving average, drawing a stem and leaf diagram, completing, and interpreting a control chart and using capture/recapture to estimate a population.” It adds a reminder on working: “Students are reminded to show any working as many students failed to do this and as a result were awarded fewer marks.”
On Paper 1 Q18(c), where students had to identify outliers by calculation, the report says: “It was not unusual to see responses involving 1.5 or 2 standard deviations rather than 3.” On Q9 it says “This question was not well-answered with over half of students failing to gain a single mark.”
The Paper 2 report lists the topics where students struggled: “Areas/topics/skills where students struggled: • describe sample frame • reasoning with histograms • suitability of comparative pie charts • appropriateness to remove outlier • interpreting regression formula • commenting on how changes in scores/weightings affect the weighted mean” It also lists where they excelled: “Areas/topics/skills where students excelled: • interquartile range • population from birth rate • calculate the probability of risk • identify multivariate data • control group – hypothesis comment • calculate skew • calculate Spearman’s Rank Correlation Coefficient” On Q10(b) it says “very few mentioned that the areas of the bars represent the frequencies” On Q13(c)(iii) it says “Nearly three-quarters failed to score with an additional 10% not responding.” On Q15 it says “More than half of the students failed to score.”
Book a free consultation to talk through Higher tier AQA GCSE Statistics tuition, whichever topics a student finds hardest.
The table lists every counted part. Basis: V is a negative verdict word, S a stated share below half with no verdict word, F few, very few or a small share. The topic label is ours.
Higher paper
Part
Topic (our label)
Marks
Basis
Column (our classification)
Reference
Paper 1
Q3
probability of a diesel car from a two-way table
1
V
Basic
D7
Paper 1
Q5(a)
comments on patterns in a time series
2
V
Not assigned
–
Paper 1
Q9
judging a dice result against fair-dice probability
3
V
Not assigned
–
Paper 1
Q10(b)
reading a cumulative frequency step polygon
3
V
Additional
C2
Paper 1
Q10(c)
probability of two chosen games
2
V
Not assigned
–
Paper 1
Q11(a)
index numbers: greatest change
1
V
Additional
E7c
Paper 1
Q12(a)
population pyramid and stratifying
1
V
Additional
B3f
Paper 1
Q12(c)
difference between two percentiles
1
V
Higher
D3a
Paper 1
Q15(a)
reasons for a binomial model
2
V
Higher
E10b
Paper 1
Q15(b)
chance of at least one rare card
3
V
Higher
E10b
Paper 1
Q18(b)(i)
which row to remove from data
1
V
Basic
B5b
Paper 1
Q18(c)
identifying outliers by calculation
5
V
Higher
D3b
Paper 1
Q18(d)(i)
comparing average mass from Normal curves
1
V
Higher
E11a
Paper 1
Q18(f)
reliability of a Normal model from skew
2
V
Higher
E5a
Paper 2
Q5(b)(ii)
reducing non-response
1
V
Basic
B5a
Paper 2
Q5(f)
comparing two frequency polygons
2
S
Additional
C2
Paper 2
Q5(g)
reliability of two data sources
1
S
Additional
B2c
Paper 2
Q6(b)
estimating from a coin-flip survey
2
V
Not assigned
–
Paper 2
Q7(b)(ii)
advantage and disadvantage of a census
2
V
Basic
B2b
Paper 2
Q7(c)(i)
naming the sample frame
1
F
Basic
B3a
Paper 2
Q7(c)(ii)
systematic sampling and shared surnames
1
S
Basic
B3d
Paper 2
Q9(a)
median by interpolation
3
V
Basic
D1a
Paper 2
Q9(b)(ii)
interpreting skew in context
2
F
Higher
E5a
Paper 2
Q10(b)
histograms with unequal class widths
1
V
Higher
C2
Paper 2
Q11(a)
why use comparative pie charts
1
F
Higher
C1b
Paper 2
Q11(c)
comparative pie charts: angle against numbers
2
S
Higher
C1b
Paper 2
Q12
comparing teams by interdecile range
2
F
Higher
D3a
Paper 2
Q13(a)(i)
reason to remove an outlier
1
V
Additional
E5b
Paper 2
Q13(a)(ii)
reason to keep an outlier
1
V
Additional
E5b
Paper 2
Q13(b)
explanatory variable
1
S
Additional
B1c
Paper 2
Q13(c)(i)
plotting a regression line
2
V
Higher
D5
Paper 2
Q13(c)(iii)
meaning of the gradient
1
S
Not assigned
–
Paper 2
Q13(c)(v)
accuracy given Pearson’s coefficient
1
S
Higher
E9b
Paper 2
Q13(e)
effect of a change on two correlation coefficients
2
F
Higher
E9c
Paper 2
Q14
weighted mean reasoning
1
V
Higher
D1a
Paper 2
Q15
conditional probability notation
4
S
Additional
E4b
Book a free consultation to talk through Higher tier AQA GCSE Statistics tuition and the topics that lose marks.
AQA’s key dates list Higher Paper 1 (8382/1H) on 7 June 2027, starting in the afternoon, and Higher Paper 2 (8382/2H) on 16 June 2027, starting in the morning. Each lasts 1 hour 45 minutes. The gap between them is 9 days (16 − 7, our calculation). Both papers assess all specification content, so any reference from sections A to E can appear on either.
Higher tier questions answered
Which grades can a student get on AQA GCSE Statistics Higher tier?
AQA's specification says a Higher tier student “will be awarded a grade within the range of 4 to 9”. A student who “just fails to achieve grade 4” is awarded an allowed grade 3, and a student below the minimum standard for that allowed grade 3 is recorded as U (unclassified) with no certificate. In the June 2026 boundary document AQA prints 35 marks for grade 4 and 27 marks for the allowed grade 3 on the Higher tier, out of 160.
Can a student sit both tiers of AQA GCSE Statistics?
AQA's specification says “Students can only be entered for foundation or higher tier not both.” It also says a student must take two question papers at the same tier and that all question papers must be taken in the same series. For Higher, Paper 1 (8382/1H) is listed for 7 June 2027 and Paper 2 (8382/2H) for 16 June 2027, each 80 marks and 1 hour 45 minutes.
What mark gave a grade 4 on the Higher tier in each series?
Out of 160 marks, AQA's Higher row gives grade 4 at 55 in June 2019, 42 in November 2021, 36 in June 2022, 40 in June 2023, 42 in June 2024, 38 in June 2025 and 35 in June 2026. Our calculation of the share is 55 ÷ 160 = 34.4% for June 2019 against 35 ÷ 160 = 21.9% for June 2026. Each figure belongs to its own series only, and none is a forecast.
Which AQA GCSE Statistics topics are on the Higher tier only?
On our count from the specification layout, 11 of the 72 content references have text only in the Higher column: Spearman's rank correlation coefficient (D6), interpreting Pearson's coefficient (E9b), the distinction between the two (E9c), experimental against theoretical values and bias (E10a), the binomial distribution (E10b), the Normal distribution (E11a, E11b), action and warning lines (E11c), standardising two samples (E11d), the Petersen capture/recapture formula (E12d) and sample means (E13b). AQA's scheme of work, a suggestion only, first places all 11 in Year two.
What did AQA's examiners say Higher students found hard?
AQA's June 2025 report on Higher Paper 1 lists drawing and interpreting a cumulative frequency step polygon and identifying outliers by calculation among the questions not answered well. Its Paper 2 report lists describing a sample frame, reasoning with histograms, comparative pie charts, removing an outlier, interpreting a regression formula and the weighted mean as struggles. On our count, 36 of the 85 parts read were described negatively, carrying 63 of the 160 marks.
How long are the Higher papers and when are they sat?
Each Higher paper is a written exam of 1 hour 45 minutes and 80 marks, worth 50% of the GCSE. AQA's key dates list Paper 1 (8382/1H) on 7 June 2027 in the afternoon and Paper 2 (8382/2H) on 16 June 2027 in the morning, 9 days apart by our calculation (16 − 7). Both papers assess all specification content, so any reference from sections A to E can appear on either.
Do Higher candidates still lose marks on foundation-column skills?
Our reading of the June 2025 Higher reports says they did. Of the 31 counted parts we could assign to one specification column, 7 sit in Basic foundation content and 9 in Additional foundation content, against 15 in Higher content only; the 5 parts left unassigned are shown in the table above. The 16 foundation-column parts carry 25 marks and the 15 Higher-column parts carry 28. AQA says all content can be assessed on Higher papers.
Talk through Higher tier AQA GCSE Statistics
Book a free consultation to talk through Higher tier AQA GCSE Statistics tuition for a Year 11 student, or read about our GCSE Statistics tuition.
Leading Tuition is rated 4.8 out of 5 on Trustpilot from 59 reviews, read on 3 October 2026.