Free cognitive ability test

What cognitive ability tests measure

“Cognitive ability” is a broad label for the reasoning that shows up across many kinds of mental tasks. Charles Spearman noticed in 1904 that people who do well on one kind of mental test tend to do well on others, and proposed a general factor, g, behind that pattern (Wikipedia: g factor). Most cognitive ability tests, whether used for hiring, school placement or IQ estimates, are built to tap that general factor through a handful of puzzle formats.

This practice set uses five of those formats. It is not a hiring test and is not modeled on any particular publisher's test. It is a sampler, so you can see which formats come easily and which need work.

The questions in this set

Number series (three questions)

A list of numbers built by one rule. Start by writing the gaps between neighbors. Equal gaps mean add the gap. Gaps that grow by a fixed step mean extend the gaps. Gaps that double suggest multiplying. Gaps that jump around usually mean two series are taking turns, so look at every other number.

Shape sequences (two questions)

A row of tiles changes step by step: a turn, an added shape, a size change, a cycling fill, a dot moving around the edge. Compare the first two tiles, confirm on the next pair, apply once more.

3x3 matrices (two questions)

Rules run across rows and down columns. Name each rule, describe the missing tile, then eliminate options that break any rule.

Figure analogies (two questions)

A changes into B; do the same to C. Inventory every change, including small shapes moving or disappearing, and watch for two changes at once.

Mental rotation (one question)

One option is the figure turned; the others are mirror images. Track one landmark and check the part beside it stays on the same side.

Why speed matters on these tests

Cognitive ability tests are usually timed, and most people do not finish every question on a hard one. That changes the strategy. On an untimed puzzle you can afford to be thorough everywhere; on a timed test you have to be thorough where it pays and quick everywhere else. In practice:

• Answer every question. Here a wrong answer and a blank both score zero, so a guess can only help.
• Find one sure rule first and use it to eliminate. Partial reasoning still narrows the field.
• Do not let one stubborn item eat the time for three easier ones.

Here each question has its own 30-second ring, so the clock is per item rather than for the whole test. When a ring empties, the question is recorded as unanswered and the next one appears.

Switching formats without losing time

A mixed test asks you to change gears every question, and the switch costs more time than people expect. Going from a number series to a shape sequence means moving from arithmetic on gaps to visual comparison of neighbors. The fix is a one-line routine per format that you start automatically: gaps for numbers, first two tiles for sequences, row one for matrices, A-to-B for analogies, one landmark for rotation. When the routine starts without thought, the switch stops costing you.

A worked number example shows how short the routine can be. For 3, 7, 15, 31: the gaps are 4, 8, 16, so each gap doubles; also each number is double the one before plus one. Either reading gives 63. Two seconds on gaps found it.

What these tests predict, and what they do not

Cognitive tests are used in hiring because, across large groups, they relate to job performance. A meta-analysis of 85 years of personnel research estimated the validity of general mental ability for predicting job performance at 0.51 (Schmidt & Hunter, 1998). A separate 2007 meta-analysis found correlations of intelligence with education of about 0.56, with occupational level about 0.45 and with income about 0.23 (Strenze, 2007).

A correlation of 0.51 is substantial for social science, but it is far from a certainty. Many people with modest test scores do excellent work, and many high scorers do not. Those numbers describe a tendency across thousands of people. They say nothing definite about you, and a ten-question practice set says even less.

Your practice score

When you finish you get a count out of ten and a plain description of that practice performance. It is not an IQ and not a prediction of any employer's result. Read it by format: three number series, two sequences, two matrices, two analogies and one rotation item. If your misses cluster in one format, that is the one to practice. The puzzle library has fifty worked puzzles per format.

How much practice helps

Quite a bit, at first. Across many studies, retaking a cognitive ability test raised scores by about a quarter of a standard deviation on average, roughly 4 IQ points, and more when the same form was reused (Hausknecht et al., 2007). Most of that comes from knowing the formats and settling into the pace. A few practice sets before a real test are well worth the time; memorizing one set of answers is not.

Related practice and the full test

If you are preparing for a specific employer test, the CCAT practice and PI cognitive assessment practice pages use similar mixes. For the difference between an IQ test and an aptitude test, read IQ test vs aptitude test. For an estimated IQ with a percentile and confidence range, take the full IQ Octopus test; it is not an employment screening tool and cannot be used as one.