The Tip of the Iceberg

Implicit vs. Explicit Testing

We all have underlying biases that we may or may not try and cover overtly. But, if psychological tests begin to target covert beliefs, can our secrets remain hidden?

Popular psychologists, such as Gaertner, Payne, and Hugenberg, conducted various studies to answer such a question. Their main focus was on racism and how, in an overt sense, we have seen a decrease in racist acts over the years. However, does this mean that racism has truly vanished?

With a disbelief in the idea that racism had disappeared, these researchers developed studies surrounding a new concept called “modern racism”, an implicit and unconscious form of racism that we may or may not be aware of. So, it hadn’t disappeared, it had transformed.

Now, these researchers understood that beliefs were not going to be represented adequately via self-reports. This method of assessment is vulnerable to conscious calculations as to what is the “right” answer and what might appear to be the “wrong” answer, thus leading to a lack of answers based on gut/intuition. Thus, Gaertner began this attempt to assess implicit attitudes. He developed an experiment where they called Liberal and Conservative party members to see how likely they were to “help a caller with his broken down car” based on his apparent race via telephone voice. Then, Payne developed a digital model that assessed how likely you were to associate a gun or a tool with Black or White based on automatic actions from underlying beliefs. As well, Hugenberg’s studies demonstrated how we have an underlying motivation to learn the emotions of our own (sometimes arbitrary!) group, but that when a facial stimulus appears too quickly to tell what group it is a part of, our biases cannot kick in, thus emotional judgment remains the same between all groups.

So, if we consistently give self-reports, this is getting at one thing. If we give IAT or some other form of implicit test, this is getting at another thing. This is the idea behind method variance, where due to the application of the same test across traits, we might get a correlation, since the test type is influencing answers for the different traits. We need to pay attention to what processes are producing these scores.

As we can see, depending on the type of test we use, we get differing results, and unfortunately, many might only reveal just the tip of the iceberg.