What does a logical reasoning test actually measure?
A logical reasoning test measures your ability to work out the rule
governing a set of information and apply it to reach the correct
conclusion. It is a test of problem-solving, pattern recognition and
decision-making, and it deliberately avoids relying on knowledge you have
learned elsewhere.
Most versions are non-verbal and multiple choice. You are shown shapes,
symbols, sequences or short scenarios, and you reach the answer by deduction
rather than by recall. Everything you need is in the question, which is what
lets an employer set the same test across candidates from very different
backgrounds.
Employers use it because logic sits underneath so much of the working day.
The same skill that finds the next figure in a sequence is used to develop
strategy, assess risk and forecast, so a logical reasoning score is treated
as a signal of how well someone will handle unfamiliar problems on the job.
What are the different types of logical reasoning test?
“Logical reasoning” is an umbrella term, and you may be asked to sit any of
five related formats. Knowing which one you face changes how you should
practise.
- Deductive reasoning gives you stated premises, usually as word
problems, and asks for the conclusion that must follow.
- Inductive reasoning shows you shapes, patterns or sequences and asks for
the most probable next step, such as the odd one out or the missing item.
- Diagrammatic reasoning presents inputs, outputs and a set of operators,
and asks which processes produce a given effect.
- Abstract reasoning is essentially inductive or diagrammatic reasoning
under another name, asking you to find the rule linking a visual sequence.
- Critical reasoning is closer to deductive reasoning but word-based,
dealing with arguments, evidence and whether a conclusion holds.
If your invitation names the format, prepare for that. If it does not,
practise across all five, because the question styles look and feel quite
different under time pressure.
How do you spot the rule in an inductive or abstract question?
Change one feature at a time and see what moves in step with it. Inductive
and abstract questions almost always combine two or three simple rules, and
the trick is to isolate them rather than stare at the whole picture.
Run through the features in turn:
- Number: are shapes being added or removed by a fixed amount each step?
- Position and rotation: is an element moving around the frame, or turning
by a set angle?
- Colour or shading: does an item cycle through, for example, white,
dotted, then filled?
- Size: does an element grow or shrink across the sequence?
Take a worked example: if the count of triangles rises by two at every step,
the sixth shape has fourteen. Confirm your rule against every frame you have
been given, not just the last two, and then apply it once. Where several
options fit part of the pattern, the correct one obeys all the rules at once.
How do you solve diagrammatic reasoning questions?
Work out what each operator does first, then trace the input through to the
output. A diagrammatic question is really a small machine: symbols act on
shapes, and your job is to learn the machine’s rules from the examples given.
A dependable method:
- Study the worked examples where an input and its output are both shown,
and note what each operator changes: it might rotate, add, remove or recolour.
- Confirm each rule on a second example before you trust it, since one
instance can be a coincidence.
- Apply the operators in order to the question’s input, one step at a
time, writing down the intermediate state.
For grids where one cell is missing, identify the rule the other cells share
first. If, say, the number of stars in a column always equals the number of
circles, the missing cell is whichever option keeps that balance. Find the
shared rule, then pick the only option that respects it.
How do you approach a deductive logical question?
Take the statements as given and accept only the conclusion they force.
Deductive logical questions are word problems, and the wrong options are
built from reasonable-sounding leaps the premises do not actually support.
Consider a worked example. All footballers are fit and healthy. All famous
sports players are footballers. The only conclusion that follows is that all
famous sports players are fit and healthy. Reversing a premise (“all fit and
healthy people are footballers”) or introducing a new fact (“all footballers
are men”) produces the distractors.
The discipline that catches them: before accepting any conclusion, point to
the premises that make it certain. If you cannot trace it back, it is not the
answer, however plausible it reads.
How do you evaluate arguments in a critical reasoning question?
Judge an argument on whether it is both relevant and directly connected to
the question, not on whether you agree with it. Critical reasoning questions
often present a yes or no answer with a supporting statement and ask whether
that statement is a strong or a weak argument.
A strong argument does two things: it bears directly on the specific question
asked, and it gives a reason that would actually change the conclusion. A weak
argument is off to one side, relies on a generalisation, or restates the
claim without supporting it.
Two habits help:
- Set your own opinion aside. The test is whether the argument works, not
whether its conclusion happens to be true.
- Check the link, not just the topic. An argument can be on the right
subject yet fail to connect to the precise point at issue, and that is what
makes it weak.
How are logical reasoning tests scored and timed?
Logical reasoning tests are scored comparatively: one mark per correct
answer, with your total set against the average results of a comparison
group. Some employers look only at your raw score against a benchmark, while
others also weigh how quickly you worked.
Timing is tight, usually around a minute a question, and the tests are built
so that few candidates finish comfortably. That shapes tactics:
- Keep moving. A minute spent on one stubborn item is a later question you
never reach, and both cost the same mark.
- Guess rather than leave blanks. Negative marking is rare, so an
eliminated-down guess beats an empty answer. Check the instructions for the
occasional exception.
Which employers and test publishers use logical reasoning tests?
Logical reasoning tests are common for managerial roles and above, and across
graduate schemes, in sectors including law, investment banking, consultancy,
human resources, customer service and market research. If a role involves
analytical or decision-making work, expect one.
Most employers license their tests, and the format varies noticeably between
publishers, which is why practising the right style matters:
- SHL offer both inductive and deductive tests, typically up to 25
minutes, in a consistent multiple-choice format.
- Kenexa emphasise abstract pattern recognition and sequence problems.
- AON (Cut-e) use a distinctive interface and short, timed
micro-assessments, so their format rewards focused preparation.
- Saville Assessment include logical elements in abstract and
diagrammatic formats, usually timed at 20 to 25 minutes, for graduate and
managerial roles.
- Talent Q run adaptive tests, where questions get harder or easier based
on your answers and there are often more options per question.
The invitation email usually names the publisher. Find out which before you
prepare, because it changes what good practice looks like.
What is the best way to prepare?
Practise every format under the clock, and review as hard as you test.
Untimed practice builds accuracy but not the speed a logical reasoning test is
really measuring.
A sequence that works:
- Take one full test untimed first, reading every worked solution. The
test on this page is built for that. The goal is to learn how each format
behaves, not to score well.
- Move to timed practice at about a minute a question, and expect an
early dip.
- Review wrong answers and lucky guesses. A guess that landed is a rule
you have not yet learned to see.
- Weight practice towards your publisher’s format and your weakest
question type, once you know both.
