One of the most important findings of modern cognitive science is that a great deal of human decision-making happens unconsciously — that we all have subtle associations and biases, buried below the level of conscious thought, that inform our actions. One of the most interesting ways to examine the biases that affect your own thinking is to use implicit association tests, such as those offered by Harvard University's Project Implicit. These free, five-minute tests can provide a revealing look at your unconscious inclinations regarding a broad variety of topics: politics, race, gender, mental health, and more. The results could surprise you.
Implicit association tests work by asking users to sort pairs of concepts together as quickly as they can. The test then uses the speed with which the users respond to measure the relative strength of their associations. For example, an implicit association test designed to measure negative attitudes about mentally ill people works by asking users to pair up words indicating either danger (such as "violent" and "aggressive") or harmlessness (such as "safe" or "gentle") with the names of mental illnesses like schizophrenia or physical ailments like multiple sclerosis. In short, the more quickly and accurately a user asked to pair words indicating dangerous behavior with mental illnesses responds, the more likely it is that she unconsciously associates mental illness with danger. (You can read more about how implicit association tests work on Project Implicit's FAQ page.)
Implicit association tests are particularly useful for sussing out people's biases regarding delicate subjects like race and gender because they don't require the user to self-report attitudes that they consciously or unconsciously consider "bad." In fact, the results of these tests often surprise the users themselves — they find that the cultural and social context they grew up in influences their unconscious associations in ways that they would never have realized.
Founded in 1998, Project Implicit has gathered enough data over its nearly 20 years of public availability to produce some interesting insights of its own. For example, Project Implicit has featured an implicit presidential popularity test about Barack Obama since his inauguration in 2009. The test measures users' implicit preferences for Obama relative to a number of past presidents by setting up head-to-head comparisons between them. (For instance, one version of this quiz might test your implicit attitude towards Obama relative solely to your implicit attitude towards John F. Kennedy.) Given that Obama is broadly considered a divisive figure in the current American political conversation with many vocal detractors on both sides of the aisle, it's surprising to see that Project Implicit's users have favored him over such historically lauded presidents as Lincoln and Jefferson:
Another especially interesting part of Project Implicit is its selection of mental health implicit association tests, which includes a number of tests designed to assess users' own unconscious feelings about their own psychology. It can be very difficult to admit it to yourself when your mental state has taken a turn for the worse, but tests that look at how closely you associate yourself with anxiety and sadness can help you take your own psychological temperature if you feel that something's not right.
Given how brief these tests are and the general difficulties associated with evaluating people's unconscious attitudes, it's no surprise that implicit association tests have some limitations. Critics of such tests have argued that they're better at evaluating the culture that individuals come from rather than the individuals themselves, that it's possible to deliberately 'game' the tests and hide your biases if you know how they work, and that the order in which pairings appear in each test might affect the results.
The researchers behind Project Implicit are generally willing to acknowledge that their tests have shortcomings (as in their FAQ) and have worked to correct for them over the years. But in spite of these challenges, broad analyses of implicit association test studies have found that they're more reflective of people's attitudes than self-reporting measures are, especially when it comes to touchy subjects. As a result, it's fairly safe to say that you could learn a lot about yourself by taking some of them.