IAT and Priming Tests explained

This post provides an overview of Implicit Bias Tests, focusing on the Implicit Association Test (IAT) and Priming Tests. It discusses the concept of implicit bias and its influence on individuals' perceptions and behaviors, highlighting the IAT as a widely-used tool for measuring these biases. The post explains how the IAT assesses the speed of associations between different concepts, including race, gender, and age, and addresses criticisms regarding its reliability and validity. Additionally, it explores priming implicit tests, which involve exposing participants to stimuli to activate specific associations and measure changes in behavior or attitudes. The post concludes by emphasizing the significance of these tests in understanding implicit biases across various disciplines.

The following video serves as an introduction to Implicit Bias Tests.

Introduction to Implicit Research Methologies

Years ago, implicit measures became very popular among academics, who used them to study biases based on race, gender, sexuality, age, and religion. They also applied them to evaluate self-esteem in clinical psychology. Nowadays, market researchers have found other uses for implicit measures, such as understanding the subconscious preferences for products, brands or even politicians. 

Explicit techniques ask a person directly for an opinion, a belief or an attitude, and they self-report explicitly. Implicit techniques, however, try to gather information without asking directly, but by measuring participant reactions related to inherent and subconscious attitudes.

Implicit methods, which do not involve explicit self-reports such as interviews or questionnaires, have gained the interest of both business researchers and scientists. Research in social cognition has made the reasons for this interest understandable by highlighting two principal limitations of explicit procedures  (Greenwald & Banaji, 1995):

• Limitation 1: Explicit methods can be strongly biased by self-presentation strategies, which alter the way people depict themselves and others in relation to themselves.

• Limitation 2: Explicit measures are limited by introspective limits. These limits can be explained by a dual-process model (Strack & Deutsch, 2004) as differences between a propositional and an associative system of information processing: 

1. 1. The propositional system, also known as System 2 (Kahneman, 2011), corresponds to explicit reasoning processes and operates consciously but slowly.

   2. The associative system, also known as System 1 (Kahneman, 2011), corresponds to the spread of activation processes and operates quickly but with limited conscious accessibility (Schnabel et al., 2008).

Implicit Response Tests (IRTs) seek to overcome the limitations of explicit self-reports, by assessing those automatic and quick associative processes. IRTs work by assessing the strength of the association between two categories, measured by the time to respond. Hence, the tests measure reaction time between two concepts. The quicker the answer, the stronger the association. 

Timing for implicit measures

One of the main questions is between which intervals of time occur responses triggered by System 1 and by System 2. 

When a stimulus occurs (0ms), our brain starts processing the information. It takes approximately 200-300ms to process and to react to it, for example, by pressing a button. Therefore, all the responses that take place in less time than 200-300ms are considered too fast and should be removed. Responses that take longer than 650-900 ms are considered contaminated by conscious decision making and are also discarded.

There is no exact window of time where System 1 occurs because some people react faster than others, and some stimuli are more difficult to process than others. Different algorithm methods are employed to filter the data. 

There are several implicit bias tests in the market, such as the Implicit Association test (IAT), Extrinsic Affective Simon Task (EAST), Semantic Priming Test, Visual Priming Test, Go/No-Go Association Task (GNAT) or the Affect Misattribution Procedure (AMP). However, Implicit Association Tests and Priming Tests are the most commonly used by both academics and businesses.

Implicit Association Test (IAT)

The first Implicit Association Test (IAT) was developed in 1998 (Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K., 1998). Years before, in 1995, two social psychology researchers, Anthony Greenwald and Mahzarin Banaji stated the idea that if memories that are not accessible to awareness can influence our actions, associations can also influence our attitudes and behavior. 

The implicit association test (IAT) is a type of implicit bias test, developed in order to prevent access to consciousness and to avoid cognitive biases, such as the social desirability bias, that can affect responses. 

Procedure:

In the IAT, the participants have to classify, as quickly as possible, a series of stimuli (words and images) under specific instructions that involves Attribute and Categories Groups.

When instructions oblige highly associated category and attribute (e.g., Category 1 + Good) to share a response key, performance is faster than when less associated category and attribute (e.g., Category 2 + Good) share a key. This performance difference implicitly measures differential association of the 2 categories with the attribute.

The quicker the performance the stronger the association. The test should not take more than 10 minutes to be completed. 

TAKE A TEST!
You can try an IAT by clicking in the popup on the bottom (bottom-right if you are using a desktop device) of this page.  If it doesn’t appear, please, refresh your browser.
Make sure you select your native language. At the end of the test, you will see your personal IAT scores along with the scores of other participants that have already taken the test.

The University of Washington, University of Virginia, Harvard University, and Yale University are involved in the Project Implicit, in which they make available a several of IAT to gather data on the unconscious bias about discriminatory behavior, stereotypes, prejudices in races and ethnicity, religion, gender, sexual orientation, etc. You can take a test on their website. Along with your personal IAT scores, they will mention possible interpretations that have a basis in research done by those universities. 

Priming tests

The priming effect is shown when the exposure to a particular stimulus affects subsequent reactions, with or without intention. Researchers Storms (1958) and Cofer (1960) have demonstrated that by presenting B words just prior to an association test, the associative strength of an A-B pair can be temporarily increased (Cramer, 1966).

Priming Tests, also known as Associative Priming tests, are a type of implicit bias tests, designed to measure the strength of the association between two stimuli (targets) and some particular attributes (primes). The targets are the two categories we want to compare (for example, two brands), and the primes are the attributes we want to test to know how associated are to each of the targets. There are different types of priming tests depending on the type of prime used. Therefore, in the Semantic Priming Tests, the prime is a word; and in the Visual Priming Tests, the prime is an image.

Procedure:

Within Priming Tests, participants are asked to react as quickly and accurately as possible to the target by classifying it into its correct category above. A prime, which can be word (Semantic Priming) or image (Visual Priming) may be briefly flashed on the screen for approximately 250ms, just before the target. The participant is told to ignore the words and to continue to focus on correctly discriminating between the two categories.

If the target is more associated with the prime, the participant will react faster. Again, the stronger the association, the quicker the answer. Like in other implicit measures, too slow and too quick responses will be removed to ensure it is the System 1 (explained above) that is operating.

Pro/Cons of IAT, Semantic and Visual Priming tests

Advantages

1. Useful for common marketing questions. Some of the most common questions in marketing are related to changes in pack, logo, distinctive assets, etc. Marketing directors seek to ensure that the new design is better than the current one on certain attributes and this type of test (especially semantic priming) fits perfectly.
2. Save time and money. This type of tests can be carried out online, allowing a considerable increase in the availability of participants around the world, and saving time and money. The participant does not have to go to a specific location or wear any sensors. The only requirements are a connection to the internet, a device (computer, tablet or smartphone), and approximately 10 minutes of concentration to accomplish the task. Online IRT's have become one of the most scalable and cost-effective neuromarketing techniques.
3. Easy to interpret results. The IAT always gives you the same kind of information: "Concept 1 is more strongly linked to Attribute than Concept 2". No ambiguity or complexity in results.

Disadvantages

1. Measure relative associations only. Implicit response test can show us which brand or pack has a stronger link to a certain attribute, but not how strong that link is. For example, they can compare the association of brand A and brand B with an attribute, or the appeal of pack A and pack B. However, sometimes we want to know the absolute strength of an association. For example: How strongly is my brand linked to this attribute? How appealing is this pack? These questions are difficult to answer with an implicit response test.
2. The more concepts you want to compare, the more tests you need. Implicit response tests compare only two concepts, but what if you want to compare 3 or 4 concepts? In that case, you will need at least 3 or 6 tests respectively. In general, if you want to compare n concepts, you will need n(n-1)/2 tests. Therefore, as the number of concepts increased, the implicit response tests became less scalable.
3. Reveals the strength of the associations, but not the reasons behind them. Implicit response tests are very useful and easy to interpret, but they do not reveal the reasons behind the results. For example, the test may show that pack A looks more modern than pack B, but not explain why. To find out the causes, implicit response tests should be combined with other implicit techniques (EEG, GSR, BVP or eye tracking) or explicit ones (surveys, interviews, focus groups... .

Recently, companies have transitioned these tests to an online format, allowing users to access them from anywhere. Below are the advantages and disadvantages of this approach.

Applications of IRTs

Research applications

As explained above, IRTs are a technique widely used in psychology and social psychology research to measure automatic responses of prejudice, stereotypes in gender, age and religion, ethnic or racial bias, etc. 

Some selected papers on general population:

• Racial bias: 

  • Wittenbrink, B., Judd, C. M., & Park, B. (1997). Evidence for racial prejudice at the implicit level and its relationship with questionnaire measures. Journal of Personality and Social Psychology, 72(2), 262–274. https://doi.org/10.1037/0022-3514.72.2.262

  • Oswald, F. L., Mitchell, G., Blanton, H., Jaccard, J., & Tetlock, P. E. (2013). Predicting ethnic and racial discrimination: A meta-analysis of IAT criterion studies. Journal of Personality and Social Psychology, 105(2), 171–192. https://doi.org/10.1037/a0032734

• Homosexuality bias:

  • R Banse, J Seise, N Zerbes, (2001) Implicit Attitudes towards Homosexuality: Reliability, Validity, and Controllability of the IAT. Journal of Experimental Psychology 48 (2), 145-160. Link

• Gender stereotypes:

  • White, M.J., White, G.B. Implicit and Explicit Occupational Gender Stereotypes. Sex Roles 55, 259–266 (2006). https://doi.org/10.1007/s11199-006-9078-z

• Mental illness stigma: 

  • Rüsch, N., Todd, A. R., Bodenhausen, G. V., Corrigan, & P. W. (2010). Do people with mental illness deserve what they get? Links between meritocratic worldviews and implicit versus explicit stigma. European Archives of Psychiatry and Clinical Neuroscience 260:617-625. Link

• General about prejudices: 

  • Laurie A. Rudman, Anthony G. Greenwald, Deborah S. Mellott, and Jordan L. K. Schwartz (1999). Measuring the Automatic Components of Prejudice: Flexibility and Generality of the Implicit Association Test. Social Cognition: Vol. 17, No. 4, pp. 437-465. https://doi.org/10.1521/soco.1999.17.4.437

Specific populations:

• Eating disorders: 

  • Amy L. Ahern, Kate M. Bennett & Marion M. Hetherington (2008) Internalization of the Ultra-Thin Ideal: Positive Implicit Associations with Underweight Fashion Models are Associated with Drive for Thinness in Young Women, Eating Disorders, 16:4, 294-307, DOI: 10.1080/10640260802115852

• Drug addiction:

  • Marhe, R., Waters, A. J., van de Wetering, B. J. M., & Franken, I. H. A. (2013). Implicit and explicit drug-related cognitions during detoxification treatment are associated with drug relapse: An ecological momentary assessment study. Journal of Consulting and Clinical Psychology, 81(1), 1–12. https://doi.org/10.1037/a0030754

• Anxiety:

  • Egloff, B., & Schmukle, S. C. (2002). Predictive validity of an implicit association test for assessing anxiety. Journal of Personality and Social Psychology, 83(6), 1441–1455. https://doi.org/10.1037/0022-3514.83.6.1441

• Alzheimer’s:

  • David P. Salmon, Arthur P. Shimamura, Nelson Butters & Stan Smith (1988) Lexical and semantic priming deficits in patients with alzheimer's disease, Journal of Clinical and Experimental Neuropsychology, 10:4, 477-494, DOI: 10.1080/01688638808408254

• Depression:

  • Gemar, M. C., Segal, Z. V., Sagrati, S., & Kennedy, S. J. (2001). Mood-induced changes on the Implicit Association Test in recovered depressed patients. Journal of Abnormal Psychology, 110(2), 282–289. https://doi.org/10.1037/0021-843X.110.2.282

• Basketball players and aggressivity: 

  • Teubel, et al (2011). Implicit but not explicit aggressiveness predicts performance outcome in basketball players. Journal of Sport Psychology, 42, 390-400. Link

Clinical psychology 

Employed to measure a patient's attitudes before and after a psychological intervention. It’s used as a tool to measure the success of therapy. 

• Mandy Grumm, Katharina Erbe, Gernot von Collani, Steffen Nestler (2008). Automatic processing of pain: The change of implicit pain associations after psychotherapy. Behaviour Research and Therapy, 46, 701-714. https://doi.org/10.1016/j.brat.2008.02.009

• Schnabel, K., & Asendorpf, J. B. (2015). Cognitive trainings reduce implicit social rejection associations. Journal of Social and Clinical Psychology, 34(5), 365–391. https://doi.org/10.1521/jscp.2015.34.3.1

• Teachman, B. A., & Woody, S. R. (2003). Automatic processing in spider phobia: Implicit fear associations over the course of treatment. Journal of Abnormal Psychology, 112(1), 100–109. https://doi.org/10.1037/0021-843X.112.1.100

Market research and political preferences (Commercial applications):

Implicit tests are employed to analyze unconscious associations of consumers or voters and to predict their behavior within market research studies, including neuromarketing. Discover here several scientific journals to find more information on neuromarketing research discoveries. 

• Attitudes towards two brands, political parties, two logos, etc. 

• • Dominika Maison, Anthony G. Greenwald, Ralph Bruin (2001). The Implicit Association Test as a measure of implicit consumer attitudes. Polish Psychological Bulletin. 32 (1): DOI://10.1066/S10012010002

  • Brunel, F.F., Tietje, B.C. and Greenwald, A.G. (2004), Is the Implicit Association Test a Valid and Valuable Measure of Implicit Consumer Social Cognition?. Journal of Consumer Psychology, 14: 385-404. doi:10.1207/s15327663jcp1404_8

• Attitudes towards two characters (politicians, celebrities, sport stars, etc).

• • Arcuri, L., Castelli, L., Galdi, S., Zogmaister, C. & Amadori, A. (2008). Predicting the vote: Implicit attitudes as predictors of the future behavior of the decided and undecided voters. Political Psychology, 29, 369–387. Link

  • Theodoridis, A. (2013). Implicit Political Identity. PS: Political Science & Politics, 46(3), 545-549. doi:10.1017/S104909651300068

  • Pérez, E.O. Explicit Evidence on the Import of Implicit Attitudes: The IAT and Immigration Policy Judgments. Polit Behav 32, 517–545 (2010). https://doi.org/10.1007/s11109-010-9115-z

• Hidden preferences

• • Friese, M., Wänke, M. and Plessner, H. (2006), Implicit consumer preferences and their influence on product choice. Psychology & Marketing, 23: 727-740. doi:10.1002/mar.20126

• • These studies can show as well how attitudes can change after exposure to a stimulus such as an advertising piece, a product, an interface, a store, etc.

The University of Washington published a document with research papers using implicit tests by topics. 

Conclusions

Psychology researchers have been using implicit measures for 40 years, especially to investigate and discuss unconscious attitudes, stereotypes, and prejudices, originally towards racial differences, and later towards gender, religion or sexual orientation. Today, IRTs are still common research tools in universities, not only to widen social research discoveries but also to polish and improve these techniques. In addition, commercial applications such as consumer research or political preferences are being addressed and employed in order to predict consumer and voter behavior.

When IRTs are used with other techniques such as EEG, GSR, BVP or eye tracking, and with declarative procedures (interviews, focus groups or surveys), it enables researchers to have a full picture of what's happening on consumer's mind, and thus, better predict their behavior. For instance, within the Bitbrain Consumer Research Lab and Human Behavior Research Lab, include these techniques, that are frequently combined with Bitbrain's EEG, GSR, BVP, or Tobii Pro eye trackers, and with declarative procedures. 

Don't forget to take the test!
Remember that you can access it by clicking in the blue pop up that appears in the bottom-left part of this page.  If it doesn’t appear, please, refresh your browser.

About the author 

Maria López  (LinkedIn, Twitter) She is the CEO and Co-founder of Bitbrain. PhD in Computer Science at University of Zaragoza and Master in Business Administration at IE Business School.

Cristina Ocejo (LinkedIn, Twitter) She holds a BSc in Psychology from the University of Barcelona and an MSc in Neuromarketing by the Autonomous University of Barcelona.

You might be interested in:

• All you need to know about neuromarketing
• Steps to design a neuromarketing project following scientific methodology
• Ethics and neuromarketing implications for market research and business practice
• What are emotions and feelings, and how to measure them?
• How to select the best Neuromarketing master? (Only Spain)
• 3 neuromarketing TED talk videos
• 15 FAQs before selecting a neuromarketing master degree program or training course
• Top 10 sources of neuromarketing news and blogs in Spanish and English

References 

• Baumeister R.F., Hutton D.G. (1987) Self-Presentation Theory: Self-Construction and Audience Pleasing. In: Mullen B., Goethals G.R. (eds) Theories of Group Behavior. Springer Series in Social Psychology. Springer, New York, NY.
• Cofer, C. N (1960). Experimental studies on the role of verbal processes in concept formation and problem-solving. Ann. N. Y. Acad. Sci., 1960, 91, 94-107. 
• Cramer, P. (1966). Mediated priming of associative responses: The effect of time-lapse and interpolated activity. Journal of Verbal Learning and Verbal Behavior, 5(2), 163–166. doi:10.1016/s0022-5371(66)80010-5 
• Segal, S. J., & Cofer, C. N. (1960). The effect of recency and recall on word association. American Psychologist, 15, 451.
• Genschow O, Demanet J, Hersche L, Brass M (2017) An empirical comparison of different implicit measures to predict consumer choice. PLoS ONE 12(8): e0183937. https://doi.org/10.1371/journal.pone.0183937
• Greenwald, et al., (2002). A Unified Theory of Implicit Attitudes, Stereotypes, Self-Esteem, and Self-Concept. Psychological Review. Vol. 109, No. 1, 3–25
• Greenwald, A.G., & Banaji, M.R. (1995). Implicit social cognition: Attitudes, self-esteem, and stereotypes. Psychological. Review, 102, 4–27.
• Kahneman, D. (2011). Thinking, fast and slow. New York: Farrar, Straus and Giroux.
• Storms, L. H. (1958). Apparent backward association: A situational effect. Journal of Experimental Psychology, 55(4), 390–395.
• Schnabel, K., et at (2008). Understanding and Using the Implicit Association Test: V. Measuring Semantic Aspects of Trait Self-Concepts. European Journal of Personality. 22: 695–706. DOI: 10.1002/per.697
• Strack, F., & Deutsch, R. (2004). Reflective and impulsive determinants of social behavior. Personality and Social Psychology. Review, 8, 220–247.
• 2012. IAT Studies Showing Validity With “Real-World” Subject Populations. [ebook] University of Washington, pp.1-8. 
• Everything you need to know about Implicit Reaction Time (IRTs). (2015, September 30). Retrieved March 27, 2020, from 
• En.wikipedia.org. 2020. Implicit Attitude