What an aptitude test measures
An aptitude test, also called a psychometric or cognitive ability test, is a standardised test of thinking: the ability to see patterns, draw conclusions and solve new problems without having seen them before. It does not measure knowledge or experience, and the answers cannot be memorised. The result is compared with a norm group, so it says how you performed relative to others who took the same test.
In recruitment it is often called a logical reasoning test, a problem-solving test or a cognitive test. The words refer to the same thing with slightly different emphasis. The most common content is matrix reasoning: a grid of three by three figures with one missing. How those tasks work is on a page of its own.
Why employers use them
The test comes early in selection, often before the first interview, and sometimes as a filter before anyone has read the application. The reason is that it is cheap to run, the same for every candidate and hard to influence with a good cover letter. Research on selection counts general problem-solving ability among the better single indicators of how someone will do in a job, and recruiters lean on that.
For the candidate it means the test decides whether you go on, not whether you get the job. A result in the middle is usually enough. What causes trouble is doing markedly worse than you can, because the format was new, the time ran out or nerves took over. That is what practice removes.
The common parts
An aptitude test can consist of one or several parts. The three that recur:
- Logical or abstract reasoning. Matrix tasks with figures. Measures pattern recognition and inference without language. This is the part that almost always appears, and the one that can be practised best. The rules that govern the tasks are collected here.
- Numerical reasoning. Number series, tables and charts. Measures the ability to calculate and interpret figures under time pressure.
- Verbal reasoning. Short passages with statements to be judged true, false or impossible to say. Measures reading comprehension and inference in words.
This page and the app are about the logical part. Many recruitments use only that.
How the test runs
The test is taken in the browser, at home, from a link in the invitation. It usually takes between ten and twenty minutes and consists of between ten and twenty-five tasks. Most modern tests are adaptive: the next task is chosen according to whether the last one was right, so the difficulty follows the candidate. That is why two people do not see the same tasks, and why a careless mistake costs more than it seems, since the next task becomes easier and counts for less.
The time limit is either per task or for the whole test. Going back is rarely possible. The instructions are shown first, sometimes with one or two practice tasks, and it is a good idea to read them calmly even if the format is familiar.
How to prepare in a few days
- Learn the format. Solve tasks until the grid and the options feel obvious. That takes an evening. Five tasks are on the start page, with answers.
- See every rule type. There are fourteen. Anyone who has seen them all does not have to discover anything new during the test, only recognise. Every rule has a solved example.
- Understand every mistake. Do not move on from a mistake until you know which rule you missed. That is the single most important part, and the reason we built answers that explain.
- Do a few timed sessions. Pace is a skill of its own. An adaptive test with a clock feels different from solving tasks at leisure.
- Take the test rested. In daytime, undisturbed, on a computer with a large screen if possible. Sleep and concentration affect the result more than an extra hour of practice the night before.
What practice can and cannot do
Practice makes a difference on matrix tasks. Anyone who knows the format does not have to learn it during the test, makes fewer careless mistakes, uses the time better and goes in calmer. That is what separates a result that matches your ability from one that falls below it.
Practice does not turn anyone into another person. An aptitude test measures something fairly stable, and nobody can promise a particular result. The level in our app is the app's own scale, from 1 to 10, and it says how hard a task you can handle in the app, not how you will do in another test. Anyone promising more than that promises too much.