7 December 2020 Luca Artesini

The Implicit Association Test (IAT), a tool for studying the relevance of the associative connections of our memory

Share on

The IAT (Implicit Association Test) is a test that allows you to objectively investigate the strength of association between an element A and an element B, precisely through what are defined as “implicit associations”. Thanks to the measurement of the reaction times that a person takes to respond positively to a match, we can in fact infer the associative strength.
Confused? Let’s take a step back and understand its enormous potential!

Brand value assessment

Our unconscious prejudices

It is now anachronistic to think that companies are only producers of objects or suppliers of services. Although the business is inevitably linked to sales, each company also sets itself some positioning objectives from a value point of view. Founding and fundamental values ​​and principles that must inevitably pass through what we can call touch-points that the company has, with the public, whether large or niche. These points of contact can be the products themselves or, why not, their packaging; it could be an advertising campaign or the company’s logo itself. Each stylistic choice, material, price or product, brings with it a small brick that composes the company’s image that produces it. Product after product, campaign after campaign, companies have inevitably shaped their image in the world. It is legitimate to think that every entrepreneur and every marketing manager wishes to have control over how this image is created and evolves, to optimize advertising campaigns, to know how the public perceives the company whose image does not necessarily coincide with the management’s one.

What is the IAT test and in what areas is it used?

How can you test something that has this nature? You can ask this explicitly, sure, but would we really get a bias-free result? Free of preconceptions and subjectivity? On an explicit level, it is plausible to think that it is easy to associate the value “Exciting” with “Ferrari” (cars). But is it for everyone? Is it more or less if compared to “Lamborghini”? Within the same “Ferrari” brand, is “Exciting” more or less strongly associated with “Speed”? It is easy to see how the approach with explicitly answered questions begins to falter, because not even we ourselves have the awareness to go down to this degree of granularity.

Measurement of implicit constructs

Hence, as sunrise is opposed to sunset, we contrast (or flank) the implicit with the explicit and we do it with a test that uses the implicit associations, the IAT (Implicit Association Test). The IAT originates in social psychology, in particular it appears for the first time as a test to understand implicit racial associations (Greenwald et al., 1998). Investigating issues of this type with explicit questions will hardly lead to honest answers while it will be much more difficult to conceal one’s “beliefs” from a test on implicit associations.

We certainly use it in lighter terms and perhaps the simplest and most effective example, although now inflated in its appeal, is the Coca Cola’ and Red Bull’s one. If I asked you if you would associate the color red with the Coca Cola brand, you would almost certainly answer yes instantly. If I asked you if you would associate the color red with the Red Bull brand instead, you would probably have more difficulty in answering yes because you would think of the color blue. Yet, “Red” really means “red”. If we measured the timing of your answers, we would then obtain an affirmative answer with a very short reaction time due to the association between Coca Cola and the color red; while for the association with Red Bull, the reaction time would be higher (implicit association) and perhaps the explicit answer could even be negative.

How does an implicit association test work?

But in detail how does an implicit association test work? In reality, few ingredients are needed. Some brands or campaigns, creativity, logos or comparable products are needed; you need a list of values ​​to be tested and finally, but certainly not least, a group of people to be tested in person or remotely. These ingredients, however, are not enough to produce valuable results. We need a team of researchers who know how to design, identify problems and potential and above all, who are able to read the results, interpreting them according to the logic of psychology and statistics. The goal is in fact to identify the results that are statistically significant. Those results, therefore, which allow us to assume, with statistical confidence, that by repeating the test, we would obtain the same result – as the Americans would say – “over and over again”. – Or at least in 95% of cases, statistical threshold currently recognized as a cut-off to define a statistically significant result.

What does an IAT test give us?

We understand how it works. But what kind of result do we get from a test of this type? Let’s pretend that a large company has three different divisions within it, three “sub-brands”, in the same sector, but with different specific market targets. This company therefore wants to understand if the public actually perceives these three segments differently and more importantly, if the perception of the same is in line with the management’s expectations. The photograph that the IAT will return to us will therefore be a comparison in which we can compare the associative strength of the different brands to a list of values ​​of interest. Results in hand, we can therefore say that a brand is more strongly associated with a value than other brands; or find that, for some values, there are no significant differences in terms of implicit associations.

Let’s take another scenario instead, with a brand having to decide between three different logo design proposals. In this case, it’s easy to think that the test result will give us a winning logo, but that’s not necessarily the case. In this case, the implicit associations can tell us which logos push towards certain values, then leaving the company the choice of adopting one or the other design depending on the strategy to be pursued.

In conclusion, the IAT test (Implicit Association Test) and its methodological variants, but not theoretical such as the one called BARTT (Brand Association Reaction Time Test), can give us a snapshot of the associative strength between the object of analysis and value assets. The combination of explicit answers (Yes vs. No) and the reaction times of implicit associations (expressed in milliseconds and measured only on affirmative answers) represent an extremely important and reliable tool, based on psychological and neuroscientific constructs validated by numerous sector studies (see for example: Maison et al., 2004; Gordon et al., 2016).

 

Bibliography

Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. (1998). Measuring individual differences in implicit cognition: the implicit association test. Journal of personality and social psychology74(6), 1464.

Maison, D., Greenwald, A. G., & Bruin, R. H. (2004). Predictive validity of the Implicit Association Test in studies of brands, consumer attitudes, and behavior. Journal of consumer psychology14(4), 405-415.

Gordon, B. S., James, J. D., & Yoshida, M. (2016). The development of brand association measures in multiple product categories: New findings and implications for goods and service brands.

Share on

Related articles:

TAG: user testing