One of these things is not like the others. One of these things doesn't belong.

Determining which one of a series of shapes is different than the others is a commonly-used reasoning test. In simple problems, like when there are five circles and a square, the answer is obvious. As more shapes are added, and other variables like colour become relevant too, the odd one out can only be identified by considering several pieces of information at the same time. This relies on a part of your cognitive repertoire known as deductive reasoning.

In this test:

So to get maximum points, get as many correct answers as you can, as fast as possible.

Performance tips:

The purpose of this test, and others, is not to necessarily improve over time, or improve due to figuring out specific answers, but rather to be a means for individuals to assess and monitor their cognitive performance over time.

The tests are randomized and dynamic, meaning that puzzles will be different each time you play. As a result, we cannot provide answers to specific puzzles.

In Odd One Out, some of the more difficult problems require paying attention to more than one feature at a time, and it can be very difficult to pick the odd shape. It is supposed to be hard! You're also compared to people in the same boat—population norms are based on people taking the test with little feedback on correct answers. You may learn a system for determining the odd shape over time, either intuitively or consciously, based on discovering whether the shape you chose was correct or incorrect. We do not have plans to add additional feedback.

Your score on this test contributes to:

The contribution of each test to each performance category is based on a "factor analysis" that looked at how tests tend to clump together when measuring a massive set of data. The results were published in Neuron in 2012 (Hampshire, Highfield, Parkin, & Owen, 2012), where this test was called "deductive reasoning." The exact contribution of each test to each performance category may change as more data is collected.

The Odd One Out task was developed as a modern variant on classical tests of fluid intelligence, such as Raven's Progressive Matrices and the Cattell's Culture Fair Intelligence Test. The Odd One Out task differs from other intelligence tests, as the problems are generated on the fly using a complex set of algorithms. Due to this on-the-fly generation, and the ability of the task to generate many tens of thousands of novel problems, the task is suitable for training reasoning abilities or taking many repeated measures, as you cannot learn the answers to specific problems by rote. Instead, you must solve a new set of problems each time you complete the test.

Deductive reasoning tests generate a characteristic pattern of activity in the back and outer surface of the frontal lobes, at the intersection between the two hemispheres, and in the middle of the parietal lobe at the back and top of the brain. A recent study of people with damage to some of these brain regions—as a result of a stroke, for example—showed that the extent of the damage was correlated with the degree to which this task was impaired (Woolgar et al., 2010).

The deductive reasoning skills that Odd One Out puts to the test are essential in everyday life. Every time you argue that something is true because of a set of facts, you are using deductive reasoning. Odd One Out is abstracted from a real-life context, but you can imagine your brain going through similar processes whenever you have to figure out a correct answer based on complex rules. Think of your tax return: based on your country's rules, which of your income sources does not belong with the others?

Your lifestyle can improve or hinder your reasoning. Physical activity does a mind good, and cardiorespiratory fitness has been linked with performance on the Odd One Out test specifically (Loprinzi & Kane, 2015). Sleep can also affect reasoning; however, some researchers (e.g. Killgore, 2010) have concluded that rule-based reasoning like Odd One Out are not as susceptible to sleep deprivation as more divergent tasks, like those that require generating creative ideas.

"Brain training" has mixed evidence, and generally brain games like these tests do not make healthy young people better at other tests or increase real-life performance (Owen et al., 2010). However, one recent study (Corbett et al., 2015) included Odd One Out as part of a battery of tasks designed to improve cognition in older people. Not only did performance improve in other cognitive tests, but real-life outcomes in the form of instrumental activities of daily living (independently performing tasks such as laundry, food preparation, and finances) improved as well. We emphasize the use of BrainLabs tests as outcomes, rather than training, but regularly playing them certainly won't hurt, and may even help, with real-life mental sharpness over the lifespan. This is especially true if you are older or have cognitive deficits.