How hard was June 2022?
This AS Mathematics set (Paper 1: Pure & Mechanics; Paper 2: Pure & Statistics; 160 marks in 3 hours) sits squarely in the middle of the difficulty range. The opening multiple-choice and 'circle your answer' items are designed to settle nerves, and a confident student should bank the first 20–30 marks of each paper quickly.
Where the marks are
Calculus is the single biggest earner. Differentiation (stationary points, tangents and normals) and integration (area under a curve) together carry roughly a quarter of the Pure marks. Coordinate geometry of circles, factorising cubics, logarithms and exponential models are all heavily represented on Paper 2. On the applied side, Mechanics (constant-acceleration kinematics, Newton's second law, the car-and-caravan tow-bar question) is worth 27 marks on Paper 1, and Statistics (sampling, the binomial distribution and a one-tailed hypothesis test) is worth 27 marks on Paper 2.
Examiner pitfalls
The command words Show that, Fully justify and Prove appear repeatedly, and this is where method marks are won or lost. Because the answer is given in 'show that' items, you must present complete, ordered working — a correct final line with gaps in the reasoning scores poorly. Classic slips here: stopping the proof that \(m^2+n^2\) is a multiple of 2 but not 4 before reaching a watertight conclusion; forgetting the second-derivative test when justifying the nature of stationary points; and dropping the \(\pm\) (or extra solutions in range) when solving the trig equation \(\cos^2 y = 10\sin y + 4\).
Strategy
Spend about 60 minutes on each Pure Section A and 30 minutes on the applied Section B. Don't over-invest in the 1-mark objective questions, and keep the AQA formulae booklet to hand for the mechanics constant-acceleration formulae and the binomial. In Statistics, lay out the hypothesis test formally — hypotheses, distribution under \(H_0\), the probability you compare against the 10% level, and a conclusion in context.
Prediction
Topics that recur almost every series and reward revision: differentiation, integration, exponential/logarithm modelling, constant-acceleration kinematics and binomial hypothesis testing. Build fluency there first.