You believe you evaluate people fairly, based on what you know about them. And yet - under time pressure, without an opportunity to deliberate - your responses are faster in some pairings than others. Concepts that your memory has linked closely are processed together more quickly. Concepts that conflict with those links take longer.

The speed difference is small. It is often invisible to introspection. But it is measurable. And it reveals associations that exist in memory regardless of what you consciously endorse.

What implicit association is

Implicit association refers to the automatic mental links between concepts that influence the speed of categorising related items, as measured by reaction time differences. When two concepts are strongly associated in memory - when one primes the other quickly - they are processed faster as a pair than concepts with weaker or contrary associations.

The Implicit Association Test (IAT) makes this measurable by comparing reaction times in congruent versus incongruent pairing tasks. If you are consistently faster when pairing one social group with positive attributes than when pairing another social group with the same positive attributes, that difference in speed reflects the implicit association - the automatic link - in your memory for those combinations.

The mystery and the reveal

Here is the mystery: people who explicitly endorse egalitarian values sometimes produce IAT results showing strong implicit associations between social groups and negative attributes. The explicit belief and the implicit association point in different directions.

Here is the reveal: they are measuring different things. Explicit attitudes are what you consciously hold and report. Implicit associations are patterns in semantic memory - links formed through repeated exposure to cultural representations, language, and social environments - that activate automatically before deliberate evaluation. The two are not necessarily the same.

Faster reaction times in IAT trials that link a social group with a negative attribute reflect an implicit bias that may not align with explicit attitudes. This is not about identifying bad people. It is about identifying the presence of automatic associations that can influence judgment and behaviour regardless of intentions.

How it works

The mechanism is spreading activation in semantic memory. When a concept is activated - a word is read, a face is seen, a category is primed - activation spreads automatically to related concepts. If "weapon" is strongly linked to a particular category in memory, that category activates faster when "weapon" appears. If the link is weak or reversed, the activation takes longer.

The strength of these associations is shaped by frequency of co-occurrence in the environment: the patterns of representation in media, education, language, and social experience. An association does not need to be endorsed consciously to be encoded. It needs only to be encountered repeatedly.

Why it matters

Understanding implicit associations helps reveal automatic links that may influence judgment in contexts like hiring, medical diagnosis, and educational evaluation, even when individuals explicitly endorse fair treatment. A hiring manager who consciously believes in equity and yet has fast implicit associations between certain group membership and negative attributes may produce evaluations that differ from their intentions - not through deliberate discrimination, but through the operation of automatic associations before deliberate evaluation has a chance to intervene.

This does not mean implicit associations determine outcomes. The relationship between IAT scores and actual discriminatory behaviour is more complex and more contextually dependent than early research suggested. The associations are one factor among many. But they are a factor that operates below the level of conscious intention.

What it is not

The IAT does not measure explicit beliefs or attitudes. Explicit beliefs are what you consciously report. The IAT measures something separate: the speed of automatic categorisation that reflects patterns in memory, not statements of belief.

A strong IAT score does not directly predict discriminatory behaviour in all contexts. The link between implicit association measures and actual behaviour is real but variable. Structured decision-making processes, motivation to be accurate, and time for deliberation all moderate whether implicit associations convert into biased decisions.

Implicit associations are not fixed. Repeated exposure to counter-stereotypic associations - in social environments, in media, through deliberate practice - can shift implicit associations over time. They are learned patterns, which means they can be updated.

Sources