QCE Maths Methods common mistakes
QCE Maths Methods · Understand the traps and how to avoid them.
Differentiating or integrating a composite function without applying the chain rule, e.g. writing d/dx sin(3x) = 3cos(3x)·? and settling on cos(3x), or d/dx e^(2x) = e^(2x).
Students learn the standard derivatives as isolated formulas and pattern-match the outer function only, treating the inner linear expression as decoration rather than a function of x whose derivative must multiply the result.
What to do instead
Before differentiating, force the habit of labelling the inner function u explicitly and writing dy/dx = dy/du × du/dx in full. Practise a mixed set where coefficients inside brackets vary (sin(x), sin(2x), sin(3x)) so the multiplier is conspicuous, and check by differentiating the answer back mentally for simple cases.
Applying the quotient or product rule with the terms in the wrong order, for example writing the derivative of u/v as (u'v − uv')/v instead of v², or dropping the minus sign in the numerator.
The formulas are memorised as symbol strings without an anchor to meaning, so the asymmetric numerator and the squared denominator are the first casualties under exam pressure.
What to do instead
Anchor the quotient rule to a rhyme or derivation from the product rule, and always write the denominator squared immediately after writing the fraction bar, before filling the numerator. Do three or four practice derivatives where the numerator terms are deliberately different sizes so a sign error produces an obviously wrong answer.
Solving trigonometric equations over a restricted domain but supplying only the principal-value solution, e.g. giving x = π/6 for sin x = 1/2 on [0, 2π] and missing 5π/6, or ignoring negative and larger solutions for tan.
Students use the inverse function key on the calculator, which returns one value, and never invoke the symmetry of the unit circle or the periodicity of the function.
What to do instead
Adopt a fixed procedure: find the principal value, then use the ASTC sign pattern or a quick sketch of the unit circle to locate the second solution in the relevant quadrant, then add multiples of the period and filter against the stated domain. Always finish by counting how many solutions the domain should contain.
Confusing transformations of the form y = f(x − h) with horizontal dilations, or applying transformations in the wrong order, so that y = 2f(3(x − 1)) + 4 is graphed with the shift applied before the dilation.
The inside and outside of the function bracket are treated as interchangeable, and the standard order of operations is misapplied to the argument rather than to the whole expression.
What to do instead
Rewrite every function in the standard form y = a f(b(x − h)) + k before graphing, and state a, b, h, k aloud with their geometric meanings. Remember that inside-bracket changes act on x and are therefore 'backwards' (a shift right for x − h), and that the horizontal scale factor is 1/b, not b.
Solving an inequality such as (x − 1)(x + 3) < 0 by treating it as an equation and giving x < 1 and x < −3, rather than the interval −3 < x < 1.
The mechanical equation-solving reflex carries over, and the sign behaviour of the quadratic between and beyond its roots is never analysed.
What to do instead
Always sketch the parabola (or sign diagram) before writing the answer, mark the roots, and read off where the graph lies below or above the axis. For rational inequalities, add a table of signs across the critical points including the vertical asymptote, and check one test point in each interval.
Omitting the constant of integration, or writing + c but then substituting limits into an indefinite integral as though it were definite.
Antidifferentiation is learned as 'reverse differentiation' with the + c as a ritual suffix, and the distinction between an antiderivative family and a signed area is not consolidated.
What to do instead
Write '+ c' as part of every indefinite integral line, and when a definite integral is required, explicitly switch notation to square brackets with limits before substituting. Practise problems that ask for both an antiderivative and a definite integral of the same integrand so the difference is visible.
Finding a stationary point and assuming it is a maximum, or classifying it by the sign of f'' at the wrong point, without evaluating the second derivative at the correct x-value.
The second-derivative test is remembered as 'negative means maximum' without the step of substituting the stationary x-value, and endpoints of a restricted domain are forgotten entirely.
What to do instead
Use a three-line template: solve f'(x) = 0 for x, evaluate f''(x) at each solution, then state the conclusion with the value. For optimisation problems on a closed interval, always also evaluate the function at the endpoints and compare, since the global extremum may not be stationary.
Using the binomial distribution when the situation is hypergeometric (sampling without replacement from a small population), or failing to state that trials are independent with constant probability.
The binomial formula is drilled heavily and becomes the default for any 'number of successes' question, and the sampling context is skimmed rather than checked against the conditions.
What to do instead
Before any probability calculation, write down n, p, and explicitly confirm the four binomial conditions: fixed number of trials, two outcomes, constant p, independence. If sampling is without replacement from a finite population, flag it and either use the hypergeometric approach or note the approximation caveat.
Interpreting a confidence interval as a probability that the true mean lies in the interval, or confusing the standard deviation of the sample with the standard error of the mean.
The interval is computed by formula without discussing what it estimates, and the two quantities share the word 'standard' and similar symbols.
What to do instead
Write the interpretation in a fixed sentence: 'We are 95% confident that the population mean lies between ...', and explicitly note that the parameter is fixed while the interval is random. Distinguish σ/√n from s by computing both for one data set and noting how each changes as n grows.
Solving log equations such as log(x) + log(x − 3) = 1 by combining to log(x(x − 3)) = 1 and then accepting a negative or out-of-domain root without checking.
The algebraic manipulation is correct, so the student stops, forgetting that the logarithm's domain restricts x before any manipulation occurs.
What to do instead
State the domain restrictions (each argument positive) as the first line of working, then solve, then test each candidate against those restrictions and discard invalid ones explicitly. Keep a habit of substituting solutions back into the original equation rather than the combined form.