One-to-One UCAT Quantitative Reasoning Tutoring

Quantitative Reasoning is the UCAT subtest that presents data in tables, charts and short passages and asks 36 questions about it in 26 minutes. Leading Tuition prepares candidates for it one to one, drilling the four arithmetic families the UCAT Consortium names and the calculator technique its own materials describe. In 2025 the mean scaled score for Quantitative Reasoning was 661, the highest of the three cognitive subtests.

Why do we treat Quantitative Reasoning as a data-handling subtest rather than a maths test?

The Consortium is unusually direct about what it is measuring. Its own description reads: "Quantitative Reasoning assesses your ability to use numerical skills to solve problems. Questions are less to do with numerical facility and more to do with problem solving." That sentence sets our teaching order. A candidate who arrives having revised percentages for a fortnight and still scores badly is almost never failing on the arithmetic; they are failing on finding the right two cells in a table inside the time available.

So the first thing we do is separate those two failures. We sit a candidate down with a timed set and record, question by question, whether the mark was lost in locating the data or in the calculation that followed. Those are different problems with different remedies, and a programme that treats them as one thing spends its hours in the wrong place. In our experience the split is heavily weighted towards data location, which is why our early sessions look more like reading practice than maths practice.

There is a second reason we push families towards Quantitative Reasoning when tuition time is limited, and it comes out of the Consortium's published statistics rather than anybody's opinion. The 2025 decile table shows Quantitative Reasoning with a far wider spread of scores than the other two cognitive subtests.

Source: UCAT Consortium, Test Statistics 2025, based on 41,354 tests. Spreads are the difference between the 1st and 9th decile figures in that table.

Read the bottom two rows together. Quantitative Reasoning carries both the highest mean, 661 against 602 for Verbal Reasoning, and the widest gap between the tenth and ninetieth percentile candidate: 300 points, half as much again as Verbal Reasoning's 200. A subtest where most candidates cluster gives a programme very little room to move a score. Quantitative Reasoning is the opposite, and that is the honest argument for putting hours into it.

What timing budget do we teach, and where is the two minutes of reserve?

Almost every published Quantitative Reasoning guide divides the section time by the question count and hands the candidate the result. The Consortium does not. Its Quantitative Reasoning Question Tutorial says: "Practise the timing: you have approx. 40 seconds per question and you need to understand the problem, find the relevant information, and then solve it." That figure is lower than the arithmetic, and the difference is worth having.

Source: question count and subtest time from the UCAT Consortium test format table; the 40-second budget from the Consortium's Quantitative Reasoning Question Tutorial. The remaining rows are our arithmetic on those two figures.

Two minutes across a 26-minute section is not a rounding error. It is roughly the time three ordinary questions take, and it is what pays for the two or three items that genuinely need a long calculation. We teach candidates to hold that reserve consciously rather than to discover at the end that they have spent it. A candidate working to 43 seconds a question has no reserve at all and is running the section on an average that no real question obeys.

The reserve also has to absorb something the arithmetic hides. The Consortium's format table shows that most Quantitative Reasoning questions "will be shown as sets of four questions connected to the same data (this may be obtained from tables, text, graphs or charts)". The first question in a set therefore costs more than the fourth, because it pays the entry cost of reading the table and working out what its columns mean. We drill in sets of four for exactly this reason: a candidate who has only practised single questions has never rehearsed the moment that actually costs them.

One more figure gets used badly. Quantitative Reasoning is preceded by a 2-minute instruction section, the longest of the four subtests, which each carry 1 minute 30 seconds apart from this one. That time is not part of the 26 minutes and cannot be carried into it, so there is nothing to be gained by rushing it. We teach candidates to spend it settling, not reading text they already know.

How do we train the on-screen calculator, and when do we train candidates not to touch it?

The Consortium states that "a basic onscreen calculator is available for Decision Making and Quantitative Reasoning. The calculator cannot perform scientific functions." What matters more for a candidate's score is a quirk the Consortium publishes on the same page and almost nobody teaches: "The calculator opens in a separate window to the test screen and will become inactive if you click off the calculator or use a keyboard shortcut. Click back on the calculator for it to work again."

That is the mechanical source of a failure we see constantly in mocks. A candidate keys in a figure, clicks back to the table to check the next one, returns to the calculator and types — and nothing registers, because the window has gone inactive. They lose the keystrokes, they lose their place in the calculation, and worse, they lose confidence in a tool they are about to need again. Across a subtest with a two-minute reserve, a handful of those exchanges is the whole margin.

So the first calculator session in a programme is not about arithmetic at all. We rehearse a fixed sequence: read the whole question and identify every figure needed, write those figures on the laminated notebook the test centre provides, then enter the calculator once and do not leave it until the answer is out. We also check that a candidate practising at home has Number Lock on, which the Consortium flags specifically, because a candidate who has trained on a keypad that silently did nothing has trained a habit rather than a skill.

The other half of the training is knowing when not to open the calculator at all. The Consortium's own tip is to "use a combination of mental arithmetic and the calculator to achieve speed and accuracy", adding that "you may not need the calculator for all questions" and that "using mental arithmetic may save you time". We build each candidate a personal threshold for that decision, because it is not the same for everyone: a candidate with quick fraction sense should be doing far more in their head than one whose mental arithmetic is reliable only in tens.

Which four number families do we drill, and what goes wrong in each?

The Consortium names the mathematical areas it expects candidates to revisit, and they are narrow enough to be taught properly rather than surveyed. Its list runs: "Percentages: Basic, % change, reverse, equivalence to decimals and fractions. Proportions and Ratios: Proportionality - both direct and inverse. Rates: Rates of flow, speed (learn the speed, distance, time triangle). Averages: Means, Modes, Medians, Ranges." The value we add is not the list. It is knowing which part of each family goes wrong under time pressure, and setting the drill against that part rather than the family as a whole.

Source: mathematical areas from the UCAT Consortium's Quantitative Reasoning Question Tutorial. The remaining columns describe our own teaching.

The percentages row deserves an example, because the Consortium supplies one that shows what "problem solving rather than numerical facility" means in practice. In its worked population question, the model answer does not calculate a percentage increase at all. It reads: "You can see that for years 2 and 3 the urban population is around twice the previous year's. So using Year 4 as the least percentage increase, divide 3426 / 3377. You should recognise that 1.0145 implies an increase of 1.45%."

Two things are being taught there and neither is arithmetic. The first is elimination by inspection — two of the four years are dismissed by eye before any figure is entered. The second is recognising a ratio as a percentage without a second step, so that 1.0145 is simply read as 1.45%. We drill that recognition until it is automatic, because a candidate who instead subtracts, divides and multiplies has spent three operations where the official answer spends one.

How do we teach candidates to read the units and drop the surplus data?

The Consortium gives two instructions on the same slide that between them describe most avoidable Quantitative Reasoning errors: "Focus on the question being asked: some information may not be required to reach an answer" and "Read the question carefully as individual words and units may be crucial in answering correctly."

Its own fuel-consumption example turns entirely on the second. The published calculation is "0.6 x 10 x 12 x 25 /100 = 18", and the model answer explains each factor: "0.6 because 10.3 - 9.7 (difference of fuel consumption between two cars in town), 10 because 5 x 2 journeys per week, 12 because this is for a 12 week term, and 25 takes into consideration the km round trip. Fuel consumption is litres / 100km so divide all by 100." A candidate who never registered that the table's unit was litres per 100 km gets an answer a hundred times too large and picks whichever option looks closest.

Notice also what the question does not use. The table it comes with carries carbon dioxide emissions for both cars and separate motorway consumption figures, and none of that is needed. So the drill we set is a unit audit performed before any keystroke: name the unit on every figure you are about to use, name the unit the answer must be in, and cross out the columns you will not touch. It sounds slow. It takes a few seconds and it removes the single largest class of error we see in mock marking.

What do we do differently for the graph questions?

Quantitative Reasoning data comes from "tables, text, graphs or charts", and the graph items behave differently enough to need separate teaching. The Consortium's own graph example is a growth chart carrying three plotted lines — a solid median and two dashed and dotted percentile lines — and asks for the average rate of increase of the median between birth and one year old.

Its model answer is a lesson in reading rather than computing: "the median line begins at 50 cm (birth) and reaches 76 cm at 12 months (one year old). The boys therefore grow 26 cm during those 12 months, an average of 26/12 = 2.17 or 2.2 cm per month." The arithmetic is trivial. The whole difficulty sits in identifying which of the three lines is the median, reading two points off it accurately, and noticing that the horizontal axis is in months while the question says one year.

Our graph drill therefore starts with the axes and the key, before the question is even read. A candidate names what each axis measures, in what unit and at what interval, and identifies every plotted series. Only then do they look at what is being asked. Candidates who work the other way round — question first, chart second — consistently read the wrong line under time pressure, and it is a habit that transfers straight out of school maths, where charts usually carry one series.

When do we teach a candidate to abandon a question?

The Consortium sanctions this explicitly: "Eliminate the more obvious incorrect answers and make an educated guess. You could consider prioritising some of the simpler/quicker questions first, and flagging others to return to later." Two published facts make that advice stronger than it looks. Quantitative Reasoning questions are "worth 1 mark each", and across the UCAT "there is no negative marking for incorrect answers".

Equal marking is the part candidates underuse. A hard question and an easy one are worth exactly the same, so a candidate who spends three questions' worth of time on one difficult item has bought one possible mark at the cost of three. With five answer options and no penalty for a wrong answer, the correct play on a question that is not coming out is to eliminate what can be eliminated, choose, flag it and move.

Saying that is easy; doing it while the clock runs is not, and it is a skill we teach rather than mention. In mock review we look at where each candidate's time actually went and set a personal abandonment trigger — a point in a question at which they stop, not a feeling. We also rehearse the flag-and-return mechanic itself, so that returning to a flagged item is not the first time they have used the review screen. The timer turning yellow when less than five minutes remain is the other cue we practise against, because for most candidates that colour change is the moment triage either happens or does not.

What does a Quantitative Reasoning programme with us look like?

Every programme starts with a timed set under real conditions, marked by strand rather than as a single score, so the first session produces a diagnosis instead of a verdict. From there the sequence is usually: calculator technique and the note-taking routine, then drill blocks against whichever of the four number families the diagnostic exposed, then full timed subtests with review focused on where the time went rather than only on what was wrong.

Tutors are matched to the candidate rather than picked from a list, and the match we care about is whether they have taught this subtest to someone with the same failure pattern. Sessions run online, which for a computer-delivered test is an advantage rather than a compromise: the candidate works on a screen, with a keypad, in the format the test uses.

We are rated Excellent on Trustpilot, 4.8 out of 5 from 54 reviews, and 91% of our students achieve their desired grades. That last figure is our practice-wide outcome measure across every level we teach. What we do not publish is a Quantitative Reasoning score uplift, because we do not hold audited before-and-after data for this subtest that we could stand behind, and a number invented to look reassuring would be worse than none. If a provider quotes you a guaranteed points gain on Quantitative Reasoning, ask what cohort it came from.

Where does Quantitative Reasoning sit in the rest of your UCAT preparation?

Quantitative Reasoning is one of three cognitive subtests alongside Decision Making and Verbal Reasoning, with Situational Judgement scored separately in bands. Abstract Reasoning was removed in 2025, which is why the Consortium warns that "scores from 2025 are not directly comparable with those from before 2025" — worth remembering if you are working from an older book.

The on-screen calculator is shared with Decision Making and no other subtest, so the calculator work described above pays twice. For the section-by-section question counts and timings across the whole test, see our UCAT tutor page; for how a full programme across all four subtests is put together, see UCAT preparation or the UCAT hub. This page deliberately does not repeat what those pages already set out.

Reviewed against the UCAT Consortium's test format and scoring, test tools, question tutorials and 2025 test statistics on 18 August 2026. As of that date the Consortium had not published mean scores or deciles for the current cycle; it states these appear in mid-September. This page describes our tuition service and is reviewed when the Consortium changes the subtest.

Frequently asked questions about UCAT Quantitative Reasoning