NCEA Numeracy common mistakes
NCEA Numeracy · Understand the traps and how to avoid them.
When converting between units, the candidate multiplies when they should divide (for example turning 350 cm into metres by writing 35,000 m, or turning 2.5 kg into grams by writing 0.0025 g).
They learn 'milli means times 1000' as a rule without attaching a direction to it, so at the moment of answering they guess whether the number should grow or shrink. Larger unit to smaller unit needs multiplication, but the mnemonic does not encode which is which, and under time pressure the guess is coin-flip quality.
What to do instead
Before computing, write the conversion as a fraction with units attached, e.g. 350 cm × (1 m / 100 cm), and cancel the unit words. Practise ten conversions in both directions each session, and after every answer sanity-check the direction: a metre is bigger than a centimetre, so the number of metres must be smaller than the number of centimetres.
The candidate computes a percentage increase and reports the size of the increase as the final answer, e.g. a $80 item increased by 15% is answered as $12 rather than $92.
The word 'increase' pulls attention to the change, and the intermediate value is the last number produced, so it feels like the answer. Candidates also conflate 'increase by 15%' with 'is 15%', which is a different question.
What to do instead
Adopt a two-line layout: line one writes 'increase = 0.15 × 80 = 12', line two writes 'new price = 80 + 12 = 92'. Circle the quantity the question actually asks for before starting. Do five mixed problems where some ask for the increase and some ask for the new total, and label which is which.
When a real-life task involves GST, a discount, then a further discount, the candidate adds the percentages and applies the sum once, treating 15% GST plus a 20% discount as a single 5% change.
Percentages are treated as fixed quantities that can be added like dollars, rather than as multipliers applied to a changing base. The candidate does not realise the second percentage operates on a different amount than the first.
What to do instead
Convert every percentage change to a multiplier and chain them: a 20% discount is ×0.80, adding 15% GST is ×1.15, so the combined factor is 0.80 × 1.15 = 0.92, a net 8% decrease. Work three examples where adding the percentages gives a visibly wrong answer and compare with the chained result, so the difference is felt rather than memorised.
The candidate reads a value off a graph or table at the wrong precision, for example stating a population of 43,000 from an axis marked only in steps of 10,000, or quoting a mean to four decimal places from data given to whole numbers.
Candidates believe more digits signals a better answer and do not connect precision to the markings or to the input data. They also read the top of a bar rather than the axis line it aligns with.
What to do instead
State the rule explicitly: read to half the smallest grid step, and round the final answer to the same precision as the least precise input. Practise by covering the numbers and predicting the value from the grid alone, then checking. For calculated statistics, write the rounding decision as a sentence before the final line.
When asked for the probability of an event, the candidate counts the favourable outcomes but divides by the number of different values rather than the total number of trials or items, e.g. answering 1/6 for rolling an even number on a die because there are six faces and one 'even' category.
The candidate collapses the sample space into categories and loses the count of individual outcomes inside each category. The distinction between 'number of outcomes' and 'number of distinct results' is never made explicit.
What to do instead
Force a full listing of the sample space before any division: write 1, 2, 3, 4, 5, 6, mark the favourable ones, then count. Only after both counts are written does the fraction get formed. Repeat with two-dice sums and with a bag of coloured counters where counts differ from colours.
The candidate uses a scale drawing or map scale in the wrong direction, measuring 6 cm on a 1:50,000 map and answering 1.2 cm or 6 km instead of 3 km.
The ratio 1:50,000 is read as a multiplication instruction without asking whether the answer should be bigger or smaller than the measurement. The candidate also forgets to convert centimetres to kilometres at the end.
What to do instead
Always write the scale as an equation with units: 1 cm on the map = 50,000 cm on the ground = 0.5 km. Then multiply the measured length by the ground value of one unit. Finish by asking 'is a real distance between two towns plausibly this size?' and reject answers that fail the plausibility check.
The candidate substitutes into a formula incorrectly, most often mishandling order of operations, e.g. computing the area of a circle as 2 × π × r or evaluating 3 + 2² as 25.
The candidate works left to right as a calculator would if brackets were omitted, and does not treat the exponent or the implied grouping in a formula as binding. Formula layout in the question often hides the grouping.
What to do instead
Rewrite each formula with explicit brackets before substituting, so πr² becomes π × (r × r) and the substitution line shows numbers in place. Use a consistent three-line method: formula, substitution, result. Check by estimating with round numbers first, so an order-of-magnitude error is caught before it is written down.
The candidate calculates a mean, median, or range for a grouped frequency table by averaging the category labels or by ignoring the frequencies, e.g. taking the median of 3, 7, 12 rather than of the underlying data.
The table presents categories as the visible objects, and the frequency column looks like a separate piece of information rather than a weight attached to each value. Candidates also confuse 'how many categories' with 'how many data points'.
What to do instead
Before calculating, add the frequency column to find n and write it down. For a mean, build an explicit value × frequency column and sum it. For a median, locate position (n+1)/2 in the cumulative frequency, not in the list of categories. Do one worked example where the two methods give different answers and explain which is correct.
The candidate answers a real-life problem with a mathematically valid but contextually impossible value, such as 4.7 buses, 12.5 people, or a negative length, and does not round to a sensible whole number.
The arithmetic is treated as the end of the task, and rounding is seen as an optional tidy-up rather than a decision driven by what the number represents. Candidates also default to standard rounding when the context demands rounding up.
What to do instead
Add a mandatory final step: write one sentence interpreting the number in the context, e.g. '4.7 buses means 5 buses are needed because a partial bus cannot be hired'. Practise distinguishing contexts that round up (buses, boxes, trips) from those that round down (full packs that fit) and those that keep decimals (money, averages).
The candidate solves a linear equation by moving terms without reversing the sign, e.g. going from 3x + 5 = 20 to 3x = 20 + 5, or dropping the negative when expanding −2(x − 3).
The operation is performed on one side only, or the inverse operation is applied with the wrong sign because the candidate is copying a pattern rather than reasoning about balance. The equals sign is not treated as a statement that both sides must stay equal.
What to do instead
Write each step as an operation applied to both sides on a new line, showing the operation in the margin, e.g. '−5 from both sides'. For expansions, draw arrows from the outside term to each inside term and write the signed product before simplifying. Substitute the final answer back into the original equation as a routine check.