Take a look at this video (QuickTime required). The screen will turn white for 1/2 second. Then a word will appear for about 1.5 seconds. Pay attention to the particular shade of gray the word is printed in. Next, a strip of five different grey squares will appear. Which square matches the color of the word?
It's a difficult task, but not impossible (we'll collect answers in a poll at the end of the post). A team led by Brian P. Meier had college student volunteers complete a similar task, and they were able to achieve 30 percent accuracy -- somewhat better than the 20 percent you would expect if they were choosing randomly. But as you might guess, the researchers weren't primarily interested in how accurate the viewers were. What they wanted to know is if the word itself influenced the students' choice of color.
In 1999 the controversial linguist George Lakoff and philosopher Mark Johnson argued that metaphors used in language can actually affect our perceptions. Indeed, metaphors of "light" and "dark" are frequently used by politicians to represent good and evil. But Lakoff and Johnson's research is typically applied only to linguistic meaning. Can it also apply to visual cognition? In 2004 Meier led a team which found that viewers were faster and more accurate categorizing positive words when the letters were white, and better categorizing negative words when the letters were black.
In a new set of experiments, Meier's team wanted to see if the process worked in reverse.
First they showed viewers 50 positive words ("love," "baby") and 50 negative words ("rude", "spider"), in random order. After viewers said whether the word was "positive" or "negative," they were shown a gray square, but they were told they were being shown one of two slightly different squares. The task was to say whether they were seeing the lighter or darker square (in fact there was only one square). Viewers were significantly more likely to say they had seen the dark square after negative words, and more likely to say they had seen the light square after positive words.
But perhaps describing the squares as "lighter" or "darker" biased the viewers, and their responses reflected not what they saw, but simply the difference in meanings of the words they had read. In a second experiment, the researchers presented the words using a display similar to the movie above. Viewers had to say whether the meaning of the word was "positive" or "negative." The words could be one of three different brightness levels, which matched exactly with one of the squares that followed. Next viewers used a mouse to click on the square that was the same color as the word they had just seen. Viewers were more likely to click on a lighter-colored square after seeing a positive word, and more likely to click on a darker-colored square after seeing a negative word.
But arguably these experiments still don't show that we automatically perceive positive and negative words in terms of light and dark. In one final experiment, positive or negative words were flashed on screen for a quarter of a second. Viewers were not required to categorize the words as positive or negative. Then one of two grey squares appeared, and viewers had just 4/10 of a second to identify the squares as light or dark. These squares, unlike the previous squares, were easily distinguishable:
So, how did the meaning of the word affect accuracy? Here are the results:
Viewers were significantly more accurate identifying the light square after seeing a positive word, and significantly more accurate identifying the dark square after seeing a negative word.
Meier's team argues that these studies demonstrate that the metaphors of light and dark are automatically associated with positive and negative words. They suggest this association may be due to the fact that humans are normally active during the day. Since we prefer the light, then light colors are naturally associated with positive emotions.
By the way, how did you do on the demo? Here's a poll to record your responses:
I'll give the correct response later today in the comments.
Meier, B.P., Robinson, M.D., Crawford, L.E., & Ahlvers, W.J. (2007). When "light" and "dark" thoughts become light and dark responses: Affect biases brightness judgments. Emotion, 7(2), 366-376.
- Log in to post comments
Hi everyone! I'm thinking of possibly doing something similar to this for my science project. I just started high school, and I am looking for a somewhat advanced version of this test, that will be done either at home, or through surveys given to around 50 people. Does anyone have an idea for how to conduct this? Two of my other fair ideas are: test foreign candy or toothpaste for lead or test tuna for mercury. Does anyone know where to buy a lead or mercury testing kit? Thanks!! Great Blog :)
I think you could do a simplified version of the test with a computer. It wouldn't take much to make easy movies like the first experiment in the study. I'd suggest reading the original study (find it at the library) and trying to replicate.
Yeah, that paper was published in Emotion because you have to shrink the scale of the graph to almost nothing to even see the effect, and you have to have a ton of data and (if I remember correctly) do the analysis within-subjects to get statistical significance. I'm just not sure effect-sizes like that mean anything in this domain.
So what's the answer? Inquiring minds want to know.
The answer is light-ish gray (second from right). I'd say the results of our one-item replication are inconclusive. It's interesting to me that the second most-popular item is the lightest gray, which makes some sense given our "positive" word.
Interesting articles as usual.
My three cents:
(1)The basis of this works seems similar to the implicit bias/attitude studies (e.g., study showing that judgements of positive words are faster when prior exposure is to White vs. Black faces)
(2) I like the environmental explanation for the results of the study. Wonder if it has any connection to the seasonal affective disorders (SAD and SAM), and more generally, the changes in mood that we have all likely experienced during sunny vs. cloudy days. To further support their hypothesis, might be interesting to see the response to positive and negative words in a room where the lighting is manipulated.
(3) I'm not sure if it was an illusion, but when I did the mini-experiment here, it seemed to me that the color of the letters in the word weren't uniform (e.g., the first letter looked darker than the last letter). Given that people's attention is usually more toward the right of a word when reading, I wonder if this explains the current data pattern from the poll. Such serial position effects have been observed in the Stroop literature.
Has anyone done any cross-linguistic studies of this? Because I don't think this really provides much evidence for anything without it. And hey, it's pretty easy to do on pretty much any college campus, provided you've got a native speaker around to do the actual administration.
Does the difference in presentation have any effect on the answer? The word is presented as gray text in a white box on a dark blue background, whereas the color samples are adjacent shades of gray directly on a dark blue background. I seem to remember another study on perception where differences in relative contrast could change the viewer's perception of how light or dark the object was (in this case, text.) Wouldn't it make more sense to eliminate this variable by presenting the word and the color choices in the same way?
So what's the answer? Inquiring minds want to know