Imagine your daily commute. Think about 2+2. Don’t picture pink elephants. These examples illustrate the automaticity of cognition. On highly familiar routes, people habitually and automatically navigate, sometimes taking a familiar route absent-mindedly even when they needed to run an errand some place else. I’d be willing to bet that most people reading this couldn’t help solve for “4,” and are now in a Dumbo-haze of pink elephants.

This kind of automaticity frees up cognitive resources, allowing us to focus on more complex or pressing matters. We automatically process things that are familiar and easy: You may not have solved 837+9184 nor pictured colorless green ideas quite as readily.

But automaticity can also lead us astray.

The Stroop effect—named after the person who discovered it—is one of the clearest demonstrations of this idea. We have covered the Stroop effect previously on several occasions.  In the classic version of this experiment (PDF here),  participants are shown color words whose ink either matches or does not match the color it names (so, red appears sometimes in red, but other times in green ink). The participant’s job is to name aloud the color of the word as quickly as possible. The Stroop effect is the difference in reaction time for congruent trials (red) compared to incongruent trials (red). Try for yourself here. Tongue-tied yet? What you likely experienced is being entirely unable to stop yourself from reading the color words, which dramatically interferes with your ability to say the colors of those words.

Even though the task focuses on the ink color rather than the meaning of the words, people cannot help processing the word. In many cases, reading the word still automatically occurs.

But what is the exact nature of the interference? A recent paper by Tillman, Eidels, and Finkbeiner in the Psychonomic Society’s journal Attention, Perception, & Psychophysics employed a unique method of collecting responses to address how the Stroop effect occurs. Specifically, the researchers wanted to know if there were different delays in responding depending on when the word was presented.

Previous work on the Stroop effect has just used mean reaction time – how long did a participant take to say the color for each trial. Tillman and colleagues changed two properties of the Stroop task, which allowed them to investigate how processing was affected at different time points.

First, the researchers equipped participants with a motion-tracking device on their finger. Participants had to respond by touching a card on the left or right of the display. Each card corresponded to a color – so, left for green and right for red, or vice versa.

Second, the researchers changed the timing of the appearance of each word relative to when the participant responded. The participant heard three beeps. Sometime between 300ms and 0ms before the third beep, the trial would appear. After the third beep, the participant then responded by moving their arm to the left or right, depending on the color.

The figure below shows the timeline:

This design allowed the researchers to see whether the Stroop effect persisted after participants had a lot of time to process the trial.

The researchers also varied whether each word needed to actually be read by the participant. If you try the Stroop task again, but blur your vision, so all you can see are the colors, the task becomes drastically easier. To prevent this escape from automaticity, the authors compared the standard Stroop task (just indicate the color of the word) to a new version, which they called a “forced-reading” Stroop task. In that version, the participant only responds with the color if the word is a color word (so, respond to red and red and green but not bed and queen and rod). Because the identity of the word is relevant to the type of response, every word in the forced-reading version needs to be read.

The figure below displays the arm trajectories—obtained via the motion-tracking device—from a single participant. The uncertainty in this participant’s responses are visible in the initially quite indeterminate trajectories from the starting point for the forced-reading version (bottom panel), compared to the more direct left- or right-ward movement in the standard version (top panel). In other words, when people were forced to read the word, their decision about the color ink was delayed because of the interference arising from reading the color word—note how the difference between the standard and forced-reading versions is particularly pronounced when color ink and word meaning are incongruent.

In addition, instead of simply measuring overall reaction time, Tillman and colleagues measured the participant’s arm velocity in the direction of the correct card at each time point. By that measure, participants did not show a Stroop effect for the standard version of the task. Participants moved their arm equally fast to respond to congruent and incongruent trials. However, participants did show a Stroop effect for the forced-reading version. Moreover, the Stroop effect was stronger when the word was presented later – the longer the participant had to process the word, the more an incorrect color word interfered.

The authors conclude that their study suggests that the magnitude of the Stroop effect is related to the probability that a person reads the word for any given trial. The longer a participant has to look at a word that they are required to read (to decide if it is a relevant word), the slower their arm moves. Plus, this effect only occurs for the forced-reading trials, not the standard trials.

What are the broader implications of this study?

The extent to which we are subject to the automaticity of our cognition is an important factor in psychological science. We are subject to implicit biases, which occur automatically, that can affect decision-making, judgments, and memories. Moreover, we are usually blind to this type of cognition – we don’t realize that we automatically process so much of the world around us.

And now you can really forget about those pink elephants.

Article focused on in this post:
Tillman, G., Eidels, A., &, Finkbeiner, M. (2016). A reach-to-touch investigation on the nature of reading in the Stroop task. Attention, Perception, & Psychophysics. DOI: 10.3758/s13414-016-1190-8.