Mathematics Specialised Level 4 common mistakes
Mathematics Specialised Level 4 · Understand the traps and how to avoid them.
When differentiating a composite function such as (3x^2+1)^5, the candidate writes 5(3x^2+1)^4 and stops, omitting the inner derivative 6x entirely.
The chain rule is memorised as 'bring the power down and reduce it by one', and the outer derivative is applied as if the bracket were a single variable. Because the outer step feels complete, the candidate never checks whether an inner function was differentiated.
What to do instead
Before differentiating, explicitly name the inner function u=g(x) and write dy/dx = dy/du * du/dx as a two-line scaffold. Practise until the scaffold is internalised, then always finish by asking 'what is the derivative of what is inside the bracket?' and writing it down.
On an indefinite integral, the candidate computes x^2/2 + 5 correctly but then writes the final answer without '+ C', or attaches C to a definite integral evaluation.
Constant-of-integration is treated as a cosmetic formality rather than a statement that the antiderivative is a family of functions. Under time pressure, the last line is written from memory of the integrand rather than from the meaning of the operation.
What to do instead
Adopt a rule that no indefinite integral line is complete until '+ C' is written, and mark every definite integral with square brackets and limits before substituting. In revision, deliberately mark the C in a contrasting colour on every worked example so the habit is visual as well as verbal.
When solving a trigonometric equation such as sin(2x) = 0.5 on the interval 0 <= x <= 360 degrees, the candidate gives only x = 15 and x = 75 and misses the solutions arising from the second quadrant or from the second revolution of 2x.
The candidate solves for the argument 2x, finds one principal value from the calculator, and then forgets that the interval for x must be doubled to give the interval for 2x before enumerating solutions. The general-solution structure is not used.
What to do instead
Always restate the required interval in terms of the argument first (here 0 <= 2x <= 720 degrees), then use the symmetry relations sin(theta) = sin(180 - theta) and periodicity to list every value of the argument inside that interval, and only then divide by 2. Sketch the sine curve over the doubled interval as a check that no crossing is missed.
When finding the inverse of a 2x2 matrix, the candidate divides by the determinant instead of multiplying by its reciprocal, or computes the determinant as ad - bc but then swaps the diagonal entries incorrectly, producing a matrix that does not satisfy A A^-1 = I.
The formula is recalled as a sequence of symbols without the accompanying check, and the cofactor sign pattern (swap the leading diagonal, negate the off-diagonal entries) is confused with the adjugate construction for larger matrices.
What to do instead
After every inverse, multiply the result by the original matrix and confirm the identity matrix appears. Write the formula as a labelled template with the determinant written outside the bracket as a fraction, and rehearse the 2x2 case until the sign pattern is automatic before attempting 3x3 inverses by cofactors.
In a proof by induction, the candidate verifies the base case and then, in the inductive step, assumes the statement for n = k and immediately writes the statement for n = k+1 as if it were given, without any algebraic link between the two.
The structure of the argument is learned as a ritual ('assume true for k, show true for k+1') rather than as a logical chain, so the candidate reproduces the labels of the proof without performing the actual manipulation that connects the assumption to the conclusion.
What to do instead
Force the inductive step to begin with the exact expression for the k+1 case and end with the target form, with every intermediate line justified. Practise writing the step as 'LHS(k+1) = ... = ... using the assumption at this point = ... = RHS(k+1)', and have each line checked for whether it uses the inductive hypothesis explicitly.
When computing the argument of a complex number in the second or third quadrant, the candidate uses arctan(y/x) directly and reports an angle in the first quadrant, giving a point in the wrong half-plane.
The calculator returns the principal value of arctan in the range -90 to 90 degrees, and the candidate treats this output as the argument without checking which quadrant the point actually lies in.
What to do instead
Plot the point on an Argand diagram before computing, and add or subtract 180 degrees (or pi) when the real part is negative. State the quadrant alongside the answer, and verify by converting back to modulus-argument form and comparing with the original Cartesian coordinates.
When finding the sum of an arithmetic or geometric series, the candidate substitutes n instead of n-1 into the formula, or uses the nth-term formula where the sum formula is required, producing an answer that is off by one term.
The symbols a, d, r, n and l are treated as interchangeable labels rather than as defined quantities, and the candidate does not distinguish between the number of terms and the index of the last term.
What to do instead
Before substituting, write down explicitly what n represents (the number of terms), list the first three terms and the last term, and count the terms for a small case to check the formula. For geometric series, confirm whether the sum is to n terms or to infinity before choosing the formula.
In a probability question involving two events, the candidate assumes independence and multiplies P(A)P(B) without justification, or applies the addition rule P(A or B) = P(A) + P(B) without subtracting P(A and B).
Independence is assumed because it makes the arithmetic easy, and the general addition rule is confused with the mutually exclusive special case. The candidate does not check whether the problem's wording or the given data supports independence.
What to do instead
Write down explicitly whether the events are stated or can be shown to be independent, mutually exclusive, or neither, and quote the general rules P(A union B) = P(A) + P(B) - P(A intersection B) and P(A intersection B) = P(A)P(B|A) every time. Test independence by checking whether P(A|B) equals P(A) using the given data.
When finding the equation of a tangent or normal to a curve at a given point, the candidate differentiates, substitutes the x-coordinate into the derivative to get the gradient, and then writes the equation using the gradient but the wrong point, or uses the gradient of the tangent for the normal without taking the negative reciprocal.
The three ingredients (point, gradient, line equation) are gathered in separate steps and the point is recalled from the question stem rather than from the evaluated function value, while the negative reciprocal is forgotten because the normal is treated as a minor variation of the tangent.
What to do instead
Adopt a fixed four-line routine: evaluate f(a) to get the point (a, f(a)); evaluate f'(a) to get the tangent gradient; for a normal take -1/f'(a) and state it explicitly; then substitute into y - y1 = m(x - x1) using the point just computed. Check the final equation by substituting the point back in.
When solving an inequality such as (x-1)/(x+2) < 3, the candidate multiplies both sides by (x+2) without considering that x+2 may be negative, and so fails to reverse the inequality, or omits the excluded value x = -2 from the discussion.
Linear equation habits carry over to inequalities, and the sign of a variable expression is not tracked. The domain restriction created by the denominator is treated as irrelevant to the solution set.
What to do instead
Never multiply an inequality by an expression whose sign is unknown. Instead move everything to one side, combine over a common denominator, form a single rational expression, and analyse its sign using a table of critical values and test intervals. Mark excluded values on the number line before stating the final solution set.