High-Activation: ...GREEN!? The Effects of Color on the Human Psyche

ABSTRACT
Our initial question was; how do colors effect mood? We narrowed our question into a workable experiment to discover the effects of six different colors on the participant's state of mind, disposition or feeling. We chose black, white, red, yellow, blue and green based on other research and how common they are in our environment. Our intention was to use their responses to various questions as an indication of their present mood. We anticipated that each of our colors would produce a specific mood, and from our tests we would be able to interpret the general reaction to each color.

INTRODUCTION
For our student generated lab, we decided to test the effects of color on the human psyche. Using six colors in a variety of tests, we expected to confirm or disprove our hypothesis that colors have a distinct impact upon the human psyche. We decided upon this project originally because color in nature was of interest to all of us and the results could be implemented in helpful ways in society.
Knowledge of the link between colors and psyche can be put to use in many practical situations. For instance, the most calming colors might be used to decorate an emergency room or prison cell, while colors that cause elation can be used in a wedding or festival. It also, for example, would be useful to know which colors soothe, excite, agitate, and depress when designing a daycare, or someoneís bedroom.
To help us understand the nature of mood, ìMoods as Sources of Stimulation: Relationships Between Personality and Desired Mood States,î by Rusting and Larsen was utilized. Their studies implement a system of measuring the activation of high to low energy words. For instance, the word ìintenseî was found to be a ìhigh activationî word, while the word ìtranquilî was monitored as a ìlow activationî word. We planned to extend this system of testing by taking the words Rusting and Larsen used and associating them with our six chosen colors. For instance, the color red might be found to be associated with the word ìintense,î thus passing on the attribute of high activation to the color as well as the word. By this practice we could measure energy levels in reference to color, and therefore draw parallels to mood.
In the experiment by Ainsworth, Simpson and Cassell they hypothesized how warm versus cool colors effect mood, with white as a neutral color. They suggest that red, a warm color will induce high arousal and activity, whereas blue will induce feelings of low arousal. However, their results did not support their hypothesis. Their conclusion is that their reason for error was in their methods, not in their actual hypothesis.
A similar experiment by Blandino and Jacobs involved more colors and more variation on the type of mood experienced. They used green, red, blue and yellow, along with white as a neutral, or control color. Their test was a blind test, so participants were not aware of what they were being tested for. Their results did suggest that colors do exert some effects on psychological states, although the effects were not as large or as general as had been expected.
An article by Taft similarly illustrates the association between specific colors, adjectives, and objects. We took this article into consideration when constructing our hypothesis, as it provided information for an educated guess.
By pulling together information from our research, we have devised a series of controlled tests using the colors red, yellow, green, blue, black, and white. We predict that the results will indicate the following:

Red surprise, intense
Yellow enthusiastic, lively
green calm, pleasant
white passive, neutral
blue tranquil
black unhappy

MATERIALS AND METHODS
For our primary experiment we used 6 study cubicles, each covered completely with one of the test colors: black, white, red, yellow, blue, or green. We used 66 subjects in this experiment, 11 for each color. The process for assigning people to cubicles was random: whichever cubicle was available was the sight of the next test. We decided that if the participants were told in advance of the purposes of the experiment, or that the experiment dealt with color at all, that this knowledge would bias the results. Therefore, each subject was told that they were taking a ìpersonality testî. They were told that this was a serious experiment and to in order ensure accurate results to please answer truthfully. The subject was then directed to a cubicle. No directions as to taking the test were given orally, because of the chance of variations in the explanation or even the tone of the test distributors voice. Directions for the test were written at the beginning of each section of the test.
The format of the test is as follows: Part I consists of a short story, approximately one paragraph in length, about a general subject. Throughout the story, there are missing words where the subject is instructed to fill in the blanks how he/she thinks the story should read. See Data Sheet: Personality Test Part I, for this section of the test.
Part II of the test is a series of written questions for the subject to answer. In order for the test to continue to be viewed as a personality test, not all questions relate to how color affects mood. For example:
1. Describe how you are feeling today in one word.
2. What is one of your strengths?
In these examples, question 1 would probably give indications of the subject's mood, hopefully affected by the color surrounding them. Question 2 is primarily used to maintain the notion that this is just a personality test. See Data Sheet: Personality Test Part II for this section of the test.
Part III of the Personality Test is composed of 8 different adjectives. The subject is instructed to rate how often on average he/she experiences each adjective daily on a 0 to 5 scale, 0 being never and 5 being constantly. See Data Sheet: Personality Test Part III for this section of the test.
This concludes the test. Each subject takes only one test with only one color theme. We considered other ways of testing the subjects, but we believe this will constitute the most unbiased results. With this number of people, 11 for each color, we believe the results will not suffer from not having each subject do all six color themes. We believe that the test should be new and fresh for each subject, with no former ideas about what
he will be doing We felt this would not be possible if he did the test for multiple colors.
We tested subjects at the same time on two different days. We used three colors for each day. Two group members would be responsible for recruiting test subjects, while the other two members were in charge of the test-taking process.
Analyzing the tests was somewhat challenging. In order to see results that could apply to our hypothesis, we did it in the following way. For Part I, we selected five blanks in order to be consistent through out our review process, that would tell us the most about the subject's mood (see Data Sheet: Personality Test Part I for selected blanks). Based on dictionary definitions we placed answers into one of the three categories, keeping them consistent for each blank and test. We then looked at the answers for these blanks for all 66 tests and scored the answers. Positive answers were given 1 point, negative answers 0 points, and neutral answers 1/2 point. Then for each test we added up the points and got a total, with a maximum of 5 points (all positive answers), and a minimum of 0 points (all negative answers). We also indexed every answer into the three categories: positive, negative, and neutral (see Data Sheet: Personality Test Part I Index). Finally, we entered the data into the Statview program to gain more information that will be discussed in our results.
In analyzing Part II of the Personality Test, we decided that only the first question, asking how the subject was feeling that day, could produce useful results. We therefore analyzed the answers to this question and compared them with other colors (see Data Sheet Personality Test Part II Statistics).
Again, in Part III of the test, we chose not to analyze all the data. We eliminated three words: stimulated, because of its mostly sexual responses; tranquil, because of its similarity to calm; and grouchy, because itís negative connotation effected its rating. This left us enthusiastic, cheerful, sluggish, annoyed, and calm. We then combined the ratings into three categories: 0-2 (low), 3 (medium), and 4-5(high). We entered this information into Statview by tallying the ratings of low, medium and high for each color (see Data Sheet: Personality Test Part III Statistics).
Because we did not want to risk inaccurate results, and because there just wasn't a great deal of additional help required to give the tests for this first experiment, we did not involve the class in this process. However, the following experiments involved the class.
In Class Experiment A, each person in the class received a paper sample of each of the 6 colors previously mentioned. There were 8 envelopes randomly placed around the room with a different mood written on each one. See Data Sheet: Class Experiment A for selected moods. These moods were chosen from a list put together by Cheryl L. Rusting and Randy J Larsen in their study called "Moods as Sources of Stimulation: Relationships Between Personality and Desired Mood States." The words are based on combinations of pleasant, unpleasant, active, and inactive meanings. The subjects are instructed to put each sample into whichever envelope they associate each color with. This experiment was used to compare which colors people associated to each mood, to actual relationships (see graphs included in results).Class Experiment B asked subjects to respond to six situations with the first color that came to mind. Each classmate responded to all six situations. See Data Sheet: Class Experiment B for the set of situations. After responding to posed situations for themselves, we split the class into 4 groups of 4 people. Each group is given a sheet with two of these same questions. Each group then collected answers to their questions from people in Peabody, Mary Lyon, McKee and Havighurst. This gave a large array of answers for all six questions, producing more accurate results. The purpose of this experiment is to show which colors people associate with certain moods (see Data Sheet: Class Experiment B Statistics).

Data Sheet: Personality Test Part I
Directions: Read the following story and fill in each blank with a word that you think fits. Please answer seriously. Make the story read as you think it should.

One day, Pat awoke at ____________ and got dressed
(time)
______________. The weather was ______________ that day, so Pat
(adverb) * (adjective)
wore a ________________. Then Pat went to the ______________ and
* (noun) (place)
had a ________________ time. Pat _______________ the whole way
*(adjective) * (verb)
home. Pat felt like _______________ after arriving back home. (verb)
Later that evening, Pat fixed a very ________________ dinner, and (adjective)
ate while watching __________________ on the television. That (T.V. show)
night Pat slept ________________. Pat had a dream involving
*(adverb)
___________________.
(noun)

Data Sheet: Personality Test Part II
Directions: Answer each question.
1. Describe how you are feeling today in one word. ________________
2. What would you say is one of your faults? ____________________
3. Would you say you are an optimist, pessimist, or a realist? ______________
4. What is one of your strengths? ____________________
5. What is your favorite color? ____________________

Data Sheet: Personality Test Part III
Directions: Rate the following words based on the 0-5 scale on how often you experience these moods on average.
0-never
1-rarely
2-occasionally
3-often
4-frequently
5-constantly

1. Enthusiastic ______
2. Sluggish ______
3. Annoyed ______
4. Cheerful ______
5. Grouchy ______
6. Calm ______
7. Tranquil ______
8. Stimulated ______

Data Sheet: Class Experiment A
Selected moods:
· intense
· enthusiastic
· lively
· happy
· calm
· tranquil
· unhappy
· passive

Data Sheet: Class Experiment B
Directions: Give the color you first think of for each situation.
1. Reading a good book
2. Scoring the winning goal.
3. Getting a 100% on the exam
4. A picnic
5. A car accident
6. Sleeping

HYPOTHESIZED RESULTS

We intend to use three different tests in hope that they will create a pattern we can draw conclusions from. In theory, we expect red to induce a highly activated mood and blue to induce an inactive but pleasant mood. Black will be unpleasant and active. Green and yellow could both be pleasant, but we suspect that yellow will also induce and active mood, versus green which will be inactive. White is a neutral color, therefore we are unsure of the mood it will elicit. Our idea is that it will be split between negative and positive moods. We will base our decision on the type of mood that they are experiencing by using several adjectives in each category. The following is a list of the adjectives we will be using and the category they fall into. These are also our hypothesized results:

Category: Adjectives: Color:

High-Activation: surprised, intense: Red
Activated-Pleasant: enthusiastic, lively: Red, Yellow
Pleasant: happy, cheerful, glad: Yellow
Unactivated-Pleasant: calm, at ease, serene: Green
Low-Activation: quiet, tranquil, still, passive: Blue, White
Unactivated-Unpleasant: dull, droopy, tired, bored: Blue
Unpleasant: unhappy, miserable, sad: Black
Activated-Unpleasant: fearful, anxious, nervous: Black, White

RESULTS

Before we began our actual experiments under controlled situations we decided that it would be useful to see what peopleís automatic responses were to specific colors. In order to accomplish this we applied our first two experiments: Class Experiment A and Class Experiment B.
After looking at the data produced by Class Experiment A we saw the following trends (see Class Experiment A Graphs):
*55% for ìCalmî chose Blue
*60% for ìEnthusiasticî chose Green
*86% for ìHappyî chose Yellow
*73% for ìIntenseî chose Red
*40% for ìLivelyî chose Yellow; however, 30% chose green
*78% for ìPassiveî chose White
*50% for ìTranquilî chose Blue; however, 30% chose Green
*80% for ìUnhappyî chose Black
In comparison to our original hypothesis, the color that we expected to be most popular for ìcalmî was green. Our test shoed the most popular was blue. This is an acceptable answer because of the similarity between the adjectives ìcalmî and ìtranquil.î For ìtranquil,î we expected the most popular result to be blue and the majority of subjects did choose blue. For the words ìhappy, intense, passive, and unhappy,î our hypothesis were also correct, stating that yellow, red, white and black respectively would be the most widely chosen colors.
In discord with our hypothesis were the words ìenthusiastic, and tranquil.î For ìenthusiasticî we expected the most popular color to be red or yellow. To our surprise the results proved that the most popular color (by a majority) was green. For ìtranquil,î most agreed with us and chose blue. However, a close second of the participants chose the color green. This could prove that there is discontinuity in the word/color associations for the word ìtranquil.î
Next we looked at Class Experiment B. Even though a different number of students answered for each of the situations, our data is still valid. This is because we were gauging the results of each situation individually. The more participants per question validated our experiment by providing us with a larger testing population to test. Here are the results from Class Experiment B (see Questions and graphs the Class Experiment B):
*33% chose Red for question 1, but a close second of 24% chose Yellow
*50% chose Green for question 2
*67% chose Red for question 3
*25% chose Yellow for question 4, but 22% chose Blue
*31% chose Blue for question 5
*42% chose Blue for question 6
From these results we realized that there were three questions that evoked a definite pattern in the answers of our participants: these included questions 2, 3, and 6. Situation 2 was: ìgoing on a picnic.î We feel that the 50% who responded green to this question were visualizing some aspect of nature, such as the usual surroundings in which a picnic occurs. For question 3, 67% responded with red to ìyou are involved in a serious car accident.î We perceive this association to the color of violence and/or blood that would maybe be involved. Finally, in question 6, 42% responded blue for the situation ìsleeping.î This can be related to the fact that blue is a low-activation color, and sleeping can definitely be categorized as a low-activation activity.
From these preliminary experiments we learned that people definitely do have preconceived notions of which colors they associate with specific situations.

Taking this into consideration, we looked at our data from Personality Test I. After scoring the data according to positivity, neutrality or negativity, we looked at the data of the scores that people received, according to the color environment they were placed in. These were our results:
*64% of those in the Black environment scored a 2.5 or lower on their test
*64% of those in the Green environment scored a 3.5 or higher on their test
*82% of those in the Blue environment scored a 4 or higher on their test
*For Red, there were two clusters of scores, the majority (55%) between 1.5 and 2.5, the other cluster at 4
*64% of those in the Yellow environment scored between a 3.5 and a 4.5 on their tests
*55% of those in the White scored either a 3 or 3.5 on their tests
This data is very interesting in helping us to discover the different activation energies of colors. We deducted that the lower the score, the lower the activation of a color. Proportionately, the higher the score, the higher the activation of the color will be.
To begin the comparison of our data to our hypothesis, we must closely examine the results for each color. The most popular score for black was a 2.5. This, along with the mean of all the scores of black, which was 2.818, help in supporting our hypothesis that black is an activated-unpleasant color. We believe that black had a negative association. Similarly the mean score for white was 3. Since white was a low activation color and tested as a neutral, the surprising result showed more activation to the color than expected.
The most active after black was red. This varied from our hypothesis that red would be a high-activation color. Even though the mean for red was only 2.909, the fact that there were two conflicting interests for red does say something about its effect on test studies. We believe that red clearly had an influence on those tested, since the two modes were at 2 and 4. In the case of Class Experiment A, students answered in a manner that reflected an overwhelming majority of ìhigh-activationî responses, by correlating the color with adjective ìintenseî. In Class Experiment B, students associated red with the situation of a car accident. This could invoke two emotional interpretations of the same word, explaining the graphs that we received. Some responded to red in a positive manner, while others responded negatively.
Following red was white, and then green. The results for green were surprising to us, and earned a higher activation-energy than we hypothesized. Although green only scored a mean of 3.136, it was originally placed in the ìunactivatedî category. This was expectant because of the natural tendencies it suggested to people, and the calming effect that we thought it would have. Both of these hypothesizes were based on the results of Class Experiment B, question 2. In discord with our hypothesis, the mode of the people tested responded to the test with a scored of 4. It is possible that people responded in this manner because of specific associations that relate to our preliminary data. In Class Experiment A, 60% of the responses for the emotion ìenthusiasticî were green. This correlated to the responses that we received for Personality Test I. Also, 30% of the answers for the word ìlivelyî were green. According to our hypothesis responses were expected to be closer to the answers we received for ìtranquilî when 30% responded with green.
The color yellow gained the second to highest activation energy. Overall the mean score for yellow was 3.227, but the mode of the subjects scored a 4 while in the yellow box. In fact more than half (7/11) of our test subjects scored above a 3.5 on this test. Our hypothesis was that yellow would have an ìactivated-positiveî effect on test studies, and according to these results our tests support our hypothesis.
Lastly is the color blue. Blue was the color that earned the highest mean score on the test with a 4.045. This was surprising to us because our hypothesis for blue was that it was a ìlow-activationî or ìunactivated-unpleasantî color. However, blue seemed to have the opposite effect on our subjects. Even though in Class Experiments A and B blue was the most popular response to both the adjectives ìcalmî and ìtranquilî and the situation ìsleeping.î Of all three low-activation words, blue still gained the highest score on Personality Test I. This could show a bias by our test subjects. All of our subjects were college students, and originally responded to blue with ìcalm,î ìtranquil,î and ìsleeping.î All of these adjectives are very desirable to college students because we would all like to feel calm or to be able to sleep. Fatigue is abundant in college life, making sleep a highly desirable activity among students. Since blue was so strongly associated with these feeling, it is our conclusion that people associate blue with a favorable color, not a color with a specific activation-energy. We think that our subjects do associate blue with low-activation activities such as sleeping, but we conducted our tests by connecting positivity with high- activation colors. Studentís answers manifested a preference for the color blue, but not to designation of its effect on mood.
There were some colors that we hypothesized to have similar effects as others. To compare these colors it was necessary for us to run an unpaired t-test and compare the p-values of the colors. Our hypothesis was that both black and red would have the effect of activated-unpleasant, and the colors blue and white would both have the effect of unactivated-pleasant on our test subjects. Since our hypothesis is that these colors would have similar effects as each other, our null hypothesis is that these colors will have the same effect as each other.
When comparing black and red, the p-value we received was .8679. This was larger than .05, thus proving that red and black did have a similar effect on some test cases. We then asked why. We feel that the reason that these two colors received a similar scoring was related to our original hypothesis: both colors at times have activated-unpleasant associations.
Although the association between black and red was in accord with our hypothesis, the correlation between white and blue was not. The p-value for the blue/white comparison was .0068. Since this was less than .05, we had to reject the null hypothesis and say that blue and white had different effects than each other on test cases. We believed this occurred because of the preferences our test cases had for each color. White turned out in a manner that we expected: close to 2.5, which illustrated its neutrality. However blue had a much higher activation energy than expected.
To sum up Personality Test I, the rating of activation energy from low to high was as follows:
Black, Red, White, Green, Yellow, Blue.
Speaking generally we can say that the scores black, white, green, and yellow received from Personality Test I support our hypothesis. For blue and red, even though they are in conflict with our hypothesis we definitely still saw a pattern in the effects that they had on peopleís moods.
The data from Personality Test: Part III of our test was placed into tables and the data sheets have been included. However, after careful consideration we decided the results should not contribute to our hypothesis. The results of the test are inconclusive due to the general nature of the test. We decide the responses to the question, ìRate how often you experience these moods on average on a scale of 0-5,î does not provide substantial information to support or negate our hypothesis. Not only were the possible answers too general but the manner in which the participants responded was not deliberate. As we monitored the test taking it was reported that 85% of the people hurriedly completed the last section, carelessly scoring a two or three. These numbers give similar neutral results for all of the colors. The thought put into this section of the test was not satisfactory, and has been disqualified as valuable information.

DISCUSSION

When we designed our experiments to test our hypothesis we attempted to
eliminate any room for error. Nevertheless, as in any experiment there were circumstances beyond our control. The people we found to take the tests were all of similar ages and had similar interests. We did test an equal amount of men as woman, however there were only
architect majors and western majors tested. We also had no way of verifying the mood
people were in previous to the tests. This could have skewed the results, as well as their
openmindedness, thus affecting their reactions to the colored environments.
There is room for improvement in the methods, even though there was careful consideration put into peoples associations with colors. Some color associations were
unavoidable. Another way of conducting the test that might have contributed more accurate results would be the choice of colors. Perhaps testing colors affects on mood might have been better with muted versus vibrant colors, or even using all muted colors. There was no way to control the effects of light on mood, and the weather could have also contributed.
Using the data that we have collected this semester from three successful test, (Class Experiments A and B, and Personality Test I) we feel that we have begun an interesting study on the effects that color can have on mood. We have learned how some colors have predetermined associations in unconscious memory, and how the activation energy of a color can effect the mood displayed by a person in that environment. It is interesting that even the environments we are in at Miami every day perhaps have been colored on purpose, consequently effecting our moods. From the yellow and pinks of Peabodyís hallways, to the subtle blues of Boydís classrooms, to the neutral grays and whites of Alumniís first year studio. All of these colored surroundings have the capability to influence the people working inside them. If we have accomplished nothing else, our group has certainly learned to appreciate our surroundings more. As Architects, Interior Designers, and lovers of nature we have all become more aware of the colors of the environments that we are submerged in everyday. Through further investigations of this topic, a greater understanding of the effects of color on mood could have a large impact. From research into color therapy to creating a marketable image for a corporation, color has become a powerful resource, carrying with it specific associations that will result in effecting the human psyche.

Next Article
Previous Article
Return to the Topic Menu

Here is a list of responses that have been posted to this Study...

• (11/12/02) [ Contributor: Riley Carson and Shannon Householder-Section: Myers ]
• 8th Grade Science Fair at McCook Central (9/25/00) [ Contributor: Suzy Webb-Section: Nicholson ]

Respond to this Submission!

IMPORTANT: For each Response, make sure the title of the response is different than previous titles shown above!

Visit the rest of the site!