GCSE Statistics past papers, marked against the real criteria

Six papers in the AQA and Edexcel formats, 2023 and 2024, with the mark scheme sitting beside every question. You type the calculation and the sentence that follows it; both are marked the way an examiner marks them, including the comment in context that decides whether correct arithmetic scores or not.

Open the papers, free, no card

What "marked against the criteria" actually means

Statistics is the subject where a pupil can get every number right and still drop marks, because a large part of the paper is not asking whether you can calculate. It is asking what the calculation says about the thing being measured. Here is a three-mark comparison in the style the boards use.

The general principle is worth stating plainly, because it holds across both boards and almost every comparison question on the paper: comparing distributions means comparing an average and comparing the spread, and the marks for each are written to require the comparison in the context of the data. "The IQR is smaller" is a fact about a subtraction. "The masses vary less" is a fact about tomatoes, and that is what is being examined.

Naming a method is not describing it

The data-collection questions have the same shape. Asked how to choose a representative sample of 60 from a school of 1200, a pupil writes "stratified sampling" and moves on. That is usually one mark of three. The rest are for the reason it is the right choice — each year group ends up in the sample in the same proportion as in the school — and for the arithmetic that follows, which on a year group of 300 gives 300 ÷ 1200 × 60 = 15 pupils.

Questionnaire questions work the same way. Spotting that a question is leading is the easy half. The marks sit in saying what the bias does to the results, and in a rewrite whose response boxes are genuinely usable: a time frame, a unit, and options that do not overlap and do not leave a respondent with nowhere to tick. The papers here include both kinds, with the marking split out so you can see which half of the answer you gave.

The diagrams have rules of their own

Statistics carries several diagrams that GCSE Maths either skims or never reaches, and each has a convention that quietly costs marks. Cumulative frequency is plotted against the upper class boundary, and the median, lower quartile and upper quartile are read across from n/2, n/4 and 3n/4 — for 80 values, from 40, 20 and 60.

Histograms are the reliable one for losing marks. When the classes have unequal widths the area of a bar is the frequency, so the height is frequency density, which is frequency divided by class width. Take a class 10 ≤ t < 20 holding 20 values and a class 20 ≤ t < 40 holding 30. The heights are 20 ÷ 10 = 2 and 30 ÷ 20 = 1.5, so the class with more data in it is drawn shorter. Read that histogram as a bar chart and every conclusion you draw from it is wrong.

Box plots are read the same way round every time: if the median line sits nearer the lower quartile, the longer tail is on the high side and the distribution is positively skewed. And a line of best fit is trustworthy inside the range of the data and not outside it, which is the difference between interpolation and extrapolation, and a question that is asked in some form on nearly every paper.

How this differs from the statistics inside GCSE Maths

Pupils taking Statistics usually take Maths too and expect a large overlap. It is smaller than it looks. Maths asks you to produce a statistic. Statistics asks you to run an enquiry: state a hypothesis, decide what data would test it and how to collect it without bias, process it, and then say what it does and does not show. That last step is the subject's whole character, and it is why so many marks turn on a sentence rather than a figure.

The content goes further too. Spearman's rank correlation coefficient, index numbers, time series and moving averages, standard deviation, and the vocabulary of sampling and bias sit in the Statistics specifications and not in the Maths one. The papers here follow the two specifications the subject is normally sat under, AQA's and Edexcel's, at Foundation and Higher tier, with a calculator allowed throughout. Across the six you will meet pie chart angles and stem-and-leaf diagrams, cumulative frequency and interpolated quartiles, estimated means from grouped tables, tree diagrams and Venn diagrams, histograms, standard deviation, index numbers, moving averages, and the sampling and questionnaire questions above — each marked point by point, not ticked or crossed as a whole.

The practical consequence is the one nobody believes until they see it on their own answer: in Statistics the arithmetic is the cheaper half of the mark, and the sentence naming what the arithmetic means about the data is the half most pupils never write.

The papers

43 questions across 6 papers, each with its own criteria. Everything here is written in the boards' own format and judged against their published criteria; the papers themselves belong to the boards and are not reproduced here.

Try it on your own answer

Open a paper, write what you would have written in the exam, including the wrong bits, and read what comes back. It is free to start and does not ask for a card.

Open GCSE Statistics

The other subjects

Twenty-three in one place, aligned to AQA, Edexcel and OCR.