Aug 28 2008

The Color Test

Published by under Uncategorized
Comments: 22

A friend of mine recently sent me this link for the “color test.”I get this or a similar link every now and then. It is a fun little neurological game, and I am always intrigued by the pop-neuro explanations that come along with it.

Go ahead – take the test, then come back….

This color test is an old neuro-psychological test known as the Stroop test, and the effect is known as the Stroop Effect after John Ridley Stroop who first published about the effect in English in 1935.   As is common, the effect was first described as a neuro-psychological phenomenon and then efforts were made to explain the effect as a consequence of neuroanatomy – how the brain is hardwired and how the different parts of the brain work together.

That is where the pop-neuro comes into play. All the sites I found that used the effect as a party game gave the same wrong explanation for the effect – that it is a demonstration of the right-brain/left-brain “struggle” or interaction.

This site explains the effect this way:

Actually, this test is checking your brains ability to separate the left brain functions and right brain functions. The problem is that the left side of your brain is trying to tell you to read the word, while at the same time, the right side of your brain is trying to tell you to identify the color!!!

This makes me wonder if the author copied this explanation from another site, or just made it up because it sounds good.

The Stroop Effect does indeed represent processing that is occurring in the brain to rectify different imperatives, but it has nothing to do with the left-brain/right-brain thing. The effect, as it is currently understood, is a matter of attention. A better paradigm for understanding this effect rather than left/right is top-down, or the hierarchical organization of the brain.

Our brains mostly function subconsciously – you don’t have to pay attention in order to breath, walk, or process visual input. At the same time the highest brain functions allow us to attend to a subset of the sensory information coming in, our internal thought, and even many of our automatic functions. You can voluntary control your breathing if you attend to it, otherwise your primitive brainstem will happily breathe on its own without your conscious control.

What is happening with the Stroop Effect is a phenomenon called interference. You are being asked to attend to the color of the “ink” of the words, and not to the words themselves. However, it is actually  impossible to turn off the reading of words once you learn how to read. The process of translating letters into words becomes so automatic that we cannot stop that part of our language center from doing it. It is thought, in fact, that most people do not even have the pathways in the brain that would enable us to inhibit that part of language function – it is literally impossible.

This is supported by experimental evidence that the Stroop Effect – which specifically refers to the delay in naming or clicking the “ink”color for color words – is always present. There are no experimental conditions that make it go away.

(As an aside, I was told by Terrence Hines that he found a small number of people who can be hypnotized to make the Stroop Effect go away – the first time this has been described – and he surmises they have a variant of hardwiring that allows this. However, I do not have the reference for this – so take it as hearsay for now.)

In fact this effect is so reliable that it has been used to catch spies. If a Russian agent is pretending to be American they can be tested to see if they exhibit the Stroop Effect for Russian words. If you cannot read a language, there will be no delay; if you can then you cannot avoid the delay.

What is happening then is that we are being confronted with conflicting stimuli – the color and the word. We then have to consciously suppress the word and respond to the color. But this takes an attentional effort and requires a finite amount of processing time that is unavoidable – hence the Stroop Effect.

What the test is really measuring (assuming literacy in the language tested) is attention, or mental effort – the highest order mental function in our brain hierarchy.

That this effect is wrongly explained on popular sites as a left-brain/right-brain conflict reflects the popularity of this concept. It is a highly overused concept, and is simply wrong in this case. Language is indeed typically encoded in the left temporal lobe (although about 5% of people are dominant for language on the right side and this is more common in left-handers). But color is not a right-brain process. Both hemispheres (the occipital lobes specifically) are able to see and process color information.

This common error also reflects the fact the we tend to explain phenomena with the explanatory models we have at hand. We are always, in fact, limited by our current repertoire of explanatory models or paradigms, because our understanding of the universe will never be complete.

The history of scientific progress reflects a tendency to explain as much as possible within current paradigms. If, however, those paradigms or insufficient eventually they will reach their limits. New modes of explanation will then have to be created – and this is the most creative and difficult aspect of science.

The most dramatic example of this of which I am aware was the crisis of classical physics that required Einstein and others to develop the new paradigms of relativity and quantum mechanics to solve.

Often, however, explanatory models are incrementally expanded, without the drama of a crisis of impasse and a bold new theory to smash old ideas. Also – sometimes the evidence suggests new modes of explanation before anyone realizes the inadequacy of the old modes. For example, Mendel discovered that heredity was digital rather than analog – meaning that some genetic traits were inherited in an all-or-none fashion (through genes), rather than an infinite blending of the traits from both parents.

The left-brain/right-brain notion is stuck in popular culture, but is mostly wrong. The real interaction among various brain structures is far more complex and interesting. But it is common for scientific concepts to trickle down to the popular culture through various imperfect media, where they are distorted and over-simplified into mostly wrong soundbites.

I have therefore come to believe, through my own experience and by observing others, that for any complex topic (whether scientific, some other academic speciality, or any other developed set of knowledge) which I have not studied from primary sources or experts, that whatever knowledge I have gleaned from the popular culture is almost certainly mostly wrong.

22 responses so far

22 thoughts on “The Color Test”

  1. cuervo says:

    Excellent post Steve. Stroop, spies and useful pop culture rule of thumb – all bases covered. Thanks.

  2. DavidCT says:

    There is nothing wrong with hypothetical explanations. They are always need as a starting point. The problem is the tendency to accept them as final explanations because they sound good or logical. The next step is to seek evidence of support or falsification. All too often no further steps are taken and proponents feel they have a vested interest in defending their hypothesis before they have any evidence at all.

    Promoting an idea that sounds good often enough in the pop culture is enough to give it its own “reality”. The idea of testing that reality is not something that is generally thought to be necessary. It is no surprise that the idea of how science works is so poorly understood in the pop world.

  3. Jim Shaver says:

    I had little trouble with the color test, scoring 100% all three times I attempted it. (The web site called me “the Master”.) I had to concentrate on the task, of course, but it wasn’t very difficult. “Turning off” my language recognition in favor of color was very easy. The slightly harder part was then turning my language center back on briefly, to read and select the correct button.

    In other words, I treated the problem as two distinct processes: first, identify the color without using language, and mentally “say” the color’s name; second, find the button with the name of the “spoken” color, and click it. If I had casually taken the test without concentrating on this procedure, I know I would have failed. However, being the Master of My Domain, I am too proud to be anything but perfect. 🙂

  4. Puppet_Master says:

    Steve ,

    I’m curious about another common idea that is being passed around about being able to tell when someone is lying or telling the truth just by where they look. Some sources even go further and say that if you look top right, it’s a visual creation, straight to the right it’s an audio creation, and bottom right it’s a kinesthetic creation. Then all those things looking to the left are real memories.

    Some examples:

    Looks kind of bogus too me, but would like to here your opinion on it.

  5. Steve Page says:

    And we only use 10% of our brain, apparently. 😉

  6. Skeptical Cat says:

    I had little trouble with the color test…

    That’s because the test wasn’t a very good demonstration of the Stroop Effect. To really notice how strong it is, you have to be timed doing the test with word/color mismatch and compare that time to just using colored shapes.

    I find the easiest way to decrease my time on the mismatched color words is to try and only focus on the first letter. It’s still noticeably slower, though.

  7. mat alford says:


    The whole looking to the left, looking to the right thing is pretty unreliable. For an excellent explanation of the science of lying, and finding out who is, try and get hold of a copy of Richard Wiseman’s book ‘Quirkology’. It’s an entertaining read, too…

    I guess the spy could always claim to be colourblind, and render the test useless….

  8. Skeptical Cat is correct, and I should have pointed this out more clearly. A formal Stroop test involves timing the response, and there is always a measured delay in response time due to the Stroop effect.

  9. Jim Shaver says:

    Steve and Skeptical Cat, you mean I’m not the Master afterall? Damnit.

  10. Dave S. says:

    I got a 63% just by clicking the button on the right as fast as I could, ignoring the words and colours altogether. Clicking the left button, not so good, 38%. So that’s the right-button/left-button “struggle” or interaction. 😉

  11. Rabscuttle says:

    I didn’t have much trouble because I just picked the button that wasn’t the same as the word. It should have more than two options.

    I am curious as to reactions would change if the buttons were solid colour instead of words.

  12. Skep says:

    I’m fascinated by these sorts of attentional issues.

    What is the seemingly related (at least vaguely related) effect of reduced comprehension caused by in appropriate highlighting of words called? I thought I read about it here on this blog and I’ve tried to look it up but I can’t come up with the right search terms..

  13. Dan Roy says:

    @ Rabscuttle & Jim Shaver

    Or, if the button words themselves displayed a similar discrepancy as the original word. You would, sort of, go through the process twice in order to get the correct answer. Or maybe the brain, once “switched” to “read” the colour would have an easier job in choosing the right coloured button?

  14. anandamide says:


    The idea that eye movements can be linked to sensory modality stems from NLP (or the roots of NLP – I think its given in one of Grinder and Bandlers books before the term was coined). As far as I know, there is very little if any reliable evidence to back it up. While Im aware its a (slightly) controversial position, NLP itself is something I would class as a pseudoscience – it certainly bears enough hallmarks to raise a few red flags for me. It´s also a subject I know a fair amount about, having spent a fair few years as a proponent. The best I can say is that, for me, it stoked a slow-burning interest in psychology and neuroscience (which I am now following up academically) – and made me more aware of pseudoscientific scams!

  15. Puppet_Master says:


    I’m surprised there isn’t a well designed study that experiments on this. I’ve heard this multiple times from independent sources. This wouldn’t even be hard at all to run an experiment.

    In fact, I do research for image compression at my school. We have a device that monitors eye movement mainly for the use of artifact detection. But, maybe I’ll do an experiment on this and see what comes about.

  16. daedalus2u says:

    There are related effects when trying to extract information from images with different spatial scales. This is the local vs. global bias in information extraction. There is considerable evidence that people with autism tend to have increased ability in being able to extract local information. Whether this increased local ability comes at the expense of decreased global ability has been controversial.

    A paper (open access) on the subject is here.

    It does appear that the increased ability to extract local information does not come at the expense of a decreased ability to extract global information. This is inconsistent with the “global processing deficit” idea of autism.

    This is a very important result in autism research.

    What is very interesting to me is how bias interferes with what is perceived. In the example Dr Novella gives in this blog, the meaning of a word interferes with the processing of the color that the word is printed in. In the example of the paper I cited, the large scale spatial relationships interfere with the processing of the small scale spatial relationships, but more so for people without autism.

    What I find especially interesting is how the biases of autism researchers have affected their perceptions of what cognitive tasks people with autism are doing. Superior local ability is called a “global deficit” without any real data showing it to be the case. There are many other examples where superior cognitive abilities are called a “deficit” because they are different from the abilities of the neurologically typically developing.

  17. Mr. Shaver, if the test called you The Master, by golly, you are The Master. But I’m biased.

    I got an 88%, then a 100%, and it called me The Master too, both times. We can both be ‘a’ Master, I suppose, but we cannot both be The Master.

  18. psamathos says:

    Yes, this test really is a terrible example of a Stroop task, since normally you want to do it as fast as possible. Having a timer bar counting down along with the possible responses printed on the screen only serve to confuse any possible Stroop effect. A real Stroop experiment would have as few variables as possible, and that means having as little on the screen as possible.

    The main problem in this test is that you can take as long as you want before responding, provided you respond within 4 seconds. Since a normal Stroop effect is in the order of hundreds of milliseconds, having such a long time to respond is counterproductive and essentially eliminates the Stroop effect. You have time to check your answer and correct yourself if it’s wrong, which will eliminate any errors that you might have made as a result of a Stroop effect.

  19. kid_pieces says:

    Heya, I had to write an extensive report for Stroop effect and there are certainly different results between sexes too. There are another way to measure such perception for those who suffer colour blindness.. Some use numbers/figures. Eg. made up a number 1 with small 2s. Others have tried with pictures.

    And true, that different cultures with different language background may be able to distinguish the colour much faster because they intend to ignore the word (English) as they wouldn’t know what it means!

    I am curious, what will happen if someone had corpus callosotomy?

    This experiment is incredibly a good start in understanding our brain and its functions. I am very interested in studying further about epilepsy and split-brain studies like Stroop effect is a very good tool to measure how far human brain works.

    Thanks for the article 🙂

  20. anandamide says:


    I’m hedging my bets in the above post – ‘As far as I know…’ :-p. I have an inkling that both the big claim (eye movements are related to ‘primary sensory modality’) and the smaller claim (eye movements are reliably related to truthfulness) have both been discredited, but am not certain how I came about this inkling and cannot reference it – it’s a bit too ‘freestanding’ for me to be confident in it. Certainly, however, the impression I got from a Bandler-led NLP seminar I attended circa 2001 was that ‘each person is unique’. But go ahead and do an experiment – who knows what you’ll find!

  21. Zytheran says:

    “However, it is actually impossible to turn off the reading of words once you learn how to read. The process of translating letters into words becomes so automatic that we cannot stop that part of our language center from doing it”

    Nah, don’t agree. With training you can turn it off.
    The Brain Training software on the Nintendo DS includes a Stroop Test and your score depends on how fast you can do this test.
    You go through a batch of 50 and it’s pretty easy to train yourself to attend to the colour and not the words and get 100% accuracy with less than one second per word. I’m still seeing the words but it simply isn’t a big a stimulus as the colour.

    Like all of these so called brain training software your brain can learn to do a specific task/test really well as opposed to a general cognitive gain. After training on the Stoop I can still read perfectly well when I need to. But my brain can also now bring my attention to the colour in priority to the words.

Leave a Reply