IB Maths AA HL: the mistakes examiners report most often

The integration itself is often fine. The marks go earlier, when a student tries to separate variables in an equation that needs an integrating factor, or does not recognise a homogeneous equation and the substitution y = vx.

Why it happens. The three types are usually taught and tested one at a time. In the exam nobody says which type it is, and the first decision is the one that was never practised.

How to fix it. Before solving anything, classify. Can it be written as f(y) dy = g(x) dx? If not, is it dy/dx + P(x) y = Q(x)? If not, is the right-hand side a function of y/x? Do mixed sets where the only task is to name the type.

On Paper 2 and Paper 3 students solve by hand what the calculator does in seconds: definite integrals for distance travelled, equations with no exact solution, complex numbers in polar form, systems of equations, statistics from a list.

Why it happens. Strong AA HL students are proud of their algebra, and most revision is done on Paper 1 style questions. The calculator is treated as a way to check an answer, not as a method.

How to fix it. Learn a fixed list of calculator routines until they are automatic: numerical solve, intersection of graphs, definite integral, numerical derivative, complex mode, simultaneous equations, distributions and inverse normal. On Paper 2 ask first whether the question wants an exact answer. If it does not, use the calculator and write down the equation or integral you solved, in standard notation, not the keystrokes.

Students know the three steps of induction but lose marks on the inductive step: the assumption for n = k is not stated as an assumption, or it is never actually used to reach n = k + 1.

Why it happens. Induction is practised on a handful of standard sums, so it becomes a template to fill in. When the statement is a divisibility result, a derivative, or something new in Paper 3, the template does not fit and the logic underneath is missing.

How to fix it. Write the assumption as a sentence, then mark the exact line where it is used. Practise on every type: series, divisibility, derivatives and De Moivre's theorem. Finish with a conclusion that names both the base case and the step.

The answer is given, so the marks are all for the argument, and the argument has gaps: steps are skipped, the result is assumed and worked backwards, or a proof by contradiction never states what is being assumed.

Why it happens. Most questions reward reaching a number. A given result changes the rules, and students keep writing as if the examiner only needs to see the last line.

How to fix it. In a "show that" question write every algebraic step and never start from the result. For contradiction, open with the assumption in words and close by naming the contradiction. Read a finished answer aloud: if a step needs the word "obviously", a line is missing.

Adding and multiplying is secure. Marks are lost when an equation has to be split into real and imaginary parts, when the roots of a polynomial with real coefficients come in conjugate pairs, and when De Moivre's theorem is needed to derive an identity or describe a transformation.

Why it happens. Complex numbers are taught as a set of separate techniques in three different forms. Exam questions move between Cartesian, polar and exponential form inside one question, and link to polynomials and trigonometry.

How to fix it. Practise choosing the form: Cartesian to add, polar or exponential to multiply, divide and take powers. For an equation that mixes z and its conjugate, write z = x + iy and equate real and imaginary parts; for zⁿ = w use polar form. When a polynomial has real coefficients, link its roots to the conjugate root theorem and to the sum and product of roots.

Students can find a scalar product but cannot plan a multi-step problem: the shortest distance from a point to a line, the reflection of a point in a plane, whether a line lies in a plane, or how planes intersect.

Why it happens. Each of these needs a picture and a plan before any calculation, and three-dimensional pictures are hard. Without a sketch the formulas are applied in the wrong order.

How to fix it. Draw a rough diagram every time, even a bad one. Learn the standard plans as short recipes: a general point on the line, a direction perpendicular to it, a scalar product set to zero. For a line in a plane, check the direction is perpendicular to the normal and that one point satisfies the plane equation.

Students can evaluate a combination but cannot decide what to count: whether order matters, how to handle a restriction, or how to split into cases without counting something twice.

Why it happens. Counting has few formulas and many ways to be wrong. It gets little teaching time because the syllabus entry is short, and the exam questions are rarely routine.

How to fix it. For every problem answer three questions in writing: does order matter, are repeats allowed, what is the restriction. Handle a restriction first, either by placing the restricted items or by counting the complement. Check a method on a tiny version of the problem that can be listed by hand.

Students differentiate repeatedly to build a series that could be written down by substituting into a standard one. With the binomial expansion for rational or negative powers they forget to factor out a constant first and ignore the range of validity.

Why it happens. The standard series are in the formula booklet, so they are not memorised or recognised. Building a series from scratch always works eventually, so it becomes the habit, and it is slow and error prone.

How to fix it. Learn to see a function as a standard series with something substituted in, or as a product of two standard series. For the binomial form, rewrite as a constant times (1 + something) to a power before expanding, and state the values of x for which it is valid. Use the first few terms to approximate a number, which is what questions usually ask next.

Given a probability density function, often piecewise, students cannot set up the integral for the median, confuse the mode with the mean, or forget to subtract the square of the mean when finding the variance.

Why it happens. This topic sits where calculus meets statistics and is often taught quickly near the end of the course. The definitions are simple, but each one is a different integral and they are easy to mix up.

How to fix it. Keep four definitions on one card: total area is 1, the median m makes the integral up to m equal to one half, the mean is the integral of x f(x), and the variance is the integral of x squared f(x) minus the mean squared. For a piecewise function, first find which piece contains the median. On Paper 2 let the calculator do the integral.

Marks are lost on the range of a function after a transformation when its domain is restricted, the domain of an inverse, the graphs of the modulus, the reciprocal and the square of a function, and sketches where no equation is given.

Why it happens. Students rely on the calculator to draw graphs and read features from the screen. When the function is not given explicitly, or the paper is non-calculator, there is nothing to type in.

How to fix it. Practise sketching from features only: intercepts, asymptotes, turning points, behaviour for large x. Remember that the domain of the inverse is the range of the original. For a transformed function, track what happens to the endpoints of the domain and to the turning points.

Simple chain rule is fine. Errors appear when it is combined with the product rule, applied to inverse trigonometric functions, written for a general function in Paper 3, or used to connect two rates of change.

Why it happens. The rule is learned as a pattern for specific functions and not as a statement about rates, so it does not transfer to a general f(g(x)) or to a geometric situation with time in it.

How to fix it. Write the chain rule in Leibniz form and say it aloud as rates: dy/dt equals dy/dx times dx/dt. In related rates, write the relationship between the variables first, differentiate with respect to time, and only then substitute numbers. With product and chain together, differentiate each factor on a separate line before combining.

Independence is confused with being mutually exclusive, the conditional probability formula is applied with the wrong event in the denominator, and Venn diagrams are filled in without using the information that two events are independent.

Why it happens. The words sound like everyday language, so students reason by intuition. The formal definitions are short and get less attention than they need.

How to fix it. Translate before calculating. Independent means P(A and B) = P(A) P(B). Mutually exclusive means P(A and B) = 0. "Given B" means B is the new whole, so P(B) goes in the denominator. Use a diagram or a table for anything with two events, and put the intersection in first.

Distance travelled is calculated as displacement, so motion that reverses direction gives the wrong answer. Maximum speed is confused with maximum velocity, and graphs of displacement against time are misread.

Why it happens. Velocity can be negative and speed cannot. The difference is one modulus sign, and it is easy to forget when the physical picture is not in mind.

How to fix it. Distance is the integral of the modulus of velocity, and on Paper 2 the calculator does it directly. For maximum speed, check the endpoints and the turning points of velocity and compare their sizes. Sketch the velocity graph before answering anything about direction.

In Section B and in Paper 3, working becomes disordered, earlier results are not reused, and information given in the question is overlooked. Students also misread what is being asked or ignore the required accuracy.

Why it happens. Long questions are designed so that each part helps the next. Under time pressure students treat every part as new and start again from nothing.

How to fix it. At each new part, ask what the previous part was for. Label results so they can be found again. Read the command term and the accuracy instruction before answering, and give answers exactly or correct to three significant figures unless the question says otherwise. Practise Paper 3 as a complete hour, not as separate parts.

Students mix up a root with a factor, do not use the sum and product of roots when coefficients are unknown, and struggle with polynomials that contain a parameter or have general degree n.

Why it happens. Polynomial work at this level is about structure, and students reach for the quadratic formula or the calculator. With unknown coefficients neither is available.

How to fix it. State the link every time: if a is a root then (x minus a) is a factor. Learn the sum and product of roots for a general polynomial and use them as the first tool when coefficients are unknown. With real coefficients, a non-real root always brings its conjugate.

Even about the x-axis the factor pi and the square are sometimes lost; about the y-axis it gets worse: students integrate with respect to the wrong variable, keep x limits where y limits are needed, or do not rearrange the curve to give x in terms of y. With a region between two curves, the volumes are subtracted incorrectly.

Why it happens. Nearly all practice is about the x-axis, so the formula is remembered as a fixed pattern in x. The formula booklet gives the y-axis version, but using it needs a rearrangement that was never rehearsed.

How to fix it. Before writing an integral, state the axis, the variable of integration and the limits in that variable. Rearrange to x as a function of y first. For a region between two curves, square each radius separately and subtract the squares, never square the difference.

The constant is dropped, added too late, or handled inconsistently when both sides of an equation are integrated. The integral of 1/x is written without the modulus, which gives wrong answers when x can be negative.

Why it happens. In definite integrals the constant cancels, so it feels optional. In differential equations and in Paper 3 it carries the boundary condition, and forgetting it changes the answer.

How to fix it. Add the constant on the line where the integration happens, once, on one side. Apply the boundary condition immediately. Write ln of the modulus every time and remove the modulus only with a reason, such as x being positive in the context.

The variance of a discrete distribution is calculated without subtracting the square of the mean, and after a linear transformation the constant is added to the variance or the multiplier is not squared.

Why it happens. Mean and variance are taught together, so the rules for one are applied to the other. The formulas are in the booklet and are not checked against meaning.

How to fix it. Remember what variance measures: spread. Adding a constant moves the distribution without changing its spread, so Var(aX + b) equals a squared times Var(X). Compute E(X squared) as its own line before subtracting the square of the mean.