The Arts in Psychotherapy 40 (2013) 115–119

Contents lists available at SciVerse ScienceDirect

The Arts in Psychotherapy

Does colour say something about emotions?: Laypersons’ assessments of colour drawings夽 Juliet Jue, PhD ATR-BC a,∗ , Seok-Man Kwon, PhD b a b

Department of Art Therapy, Han-Yang Cyber University, 222 Wangshimri-ro, Sungdong-gu, Seoul 133-791, Republic of Korea Department of Psychology, Seoul National University, 1 Kwanak-ro, Kwanak-gu, Seoul 151-742, Republic of Korea

a r t i c l e Keywords: Colour Drawing test Chromatic Achromatic

i n f o

a b s t r a c t This study examines the use of colours in estimating psychological states. Sixteen artworks were produced by two different populations: clients with a psychological condition and psychologically healthy people. Their drawings were presented in pairs on screen either chromatically or achromatically. Two hundred and twenty-four subjects were asked to choose a picture which they thought had been produced by a psychologically unhealthy client. The rate of correct answers was higher for the chromatic condition. When artworks were presented achromatically, errors increased. Analysis of colour factors and postinquiry indicated that the number of colours used, list of colours, and colour connotation might have affected choices. This study confirms that colour can be effective in estimating psychological states. Its implications and limitations are discussed and suggestions made for future study. © 2012 Elsevier Inc. All rights reserved.

Introduction Does colour say something about emotions? If messages conveyed by colours contradicted messages from other graphic indicators, what would happen? The purpose of this paper is to answer these questions using artworks and laypersons’ judgement. It has been argued that ‘colour is a subtle stimulation with salient impact’ (Jalil, Yunus, & Said, 2012). Since colour affects human lives in various ways, it is no wonder that diverse disciplines have pursued colour studies in various ways. Research on the relationship between colour and emotions has a long history. First, the psychophysical approach examines colour and human responses by means of coloured lighting. Second, psychometric research investigates colour experiences using self-report questionnaires and performance assessment. Third, clinical and experiential methods provide structured drawing tests and results. Drawing tests using coloured materials were largely developed in the late-twentieth century. They include the Diagnostic Drawing Series (Cohen, Hammer, & Singer, 1988; Cohen, Mills, & Kijak, 1994; Neale, 1994), Ulman Personality Assessment Procedure (Ulman, 1992), Person Picking an Apple from Tree (Lande, Howie, & Chang, 1997; Gantt, 2001), and the Face Stimulus Assessment (Betts, 2003). Along with the development of drawing tests, rating systems were devised to investigate graphic factors. They

夽 This article is a part of Juliet Jue’s doctoral dissertation of clinical psychology, Seoul National University in Korea and it was modified and supplemented. ∗ Corresponding author. Tel.: +82 10 2840 0903; fax: +82 2 2290 0601. E-mail address: [email protected] (J. Jue). 0197-4556/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.aip.2012.12.005

include the DDS rating system (Cohen, 1986; Cohen et al., 1994), the Descriptive Assessment of Psychiatric Artwork (Hacking, 1999), and the Formal Elements Art Therapy Scale (Gantt & Tabone, 1998). These rating systems developed colour factors which showed their relationship with psychological states. First, the number of colours used in a drawing is closely related to emotional states. Depressed people, abused victims, and chronic schizophrenic patients use a limited number of colours (Cohen et al., 1994; Gulbro-Leavitt & Schimmel, 1991; Malchiodi, 1997; Pianetti, Palacios, & Elliott, 1964; Wadeson, 1971). Second, the prominence of colour can be an important pointer to emotion (Gantt & Tabone, 1998; Gantt, 2001; Oster & Gould, 1987; Wadeson, 1971). Gantt and Tabone (1998) measured how much colour a person uses in an entire picture, and found that depressed people use colour only to outline the forms in pictures. Third, proper usage of colour and colour fitness can tell emotional appropriateness (Amos, 1982; Gantt, 2001; Malchiodi, 1997; Yokota, Ito, & Shimizu, 1999). Colour fitness assesses whether colours in drawings are appropriately used on concrete objects. Inappropriate colour usage reflects apathy of schizophrenic patients (Yokota et al., 1999), history of abuse (Malchiodi, 1997), and difficulties in emotional integration (Amos, 1982). Lastly, the list of colours and their meanings can helpful for assessing emotions (Birren, 1988; Hacking, 1999; Suenaga, 1998). Colour is clearly a compelling factor for assessing emotional states. There are issues with colour research, however. First, factor-based analysis could be a reductionist approach. There is consensus among clinicians and researchers who use drawing tests that art is a Gestalt, which is not a simple addition of parts. Therefore a single element might offer meaningful interpretations,

116

J. Jue, S.-M. Kwon / The Arts in Psychotherapy 40 (2013) 115–119

but the interpretation of the whole object could differ. Moreover, it is possible that each colour component might lead to different interpretations. Is it possible to demonstrate the effect of colour as a whole? Second, some colour research excluded other graphic factors such as shape, texture, and size to remove the compounding effect of those variables, and coloured lighting, coloured plates, and coloured environments were used in laboratory settings. These experimental designs were well controlled but the results had limited application. Is it possible to retain the graphic ingredients and only change the colour(s)? Thus, this study aimed to shed some light on the topic by using real artwork produced by different populations, and by using computer technologies to produce monochrome pictures. Monochrome pictures do not change any graphic components except colour. With these conditions in mind, we designed a forced-choice task to evaluate drawings. In this way, we sought to demonstrate phenomenologically the effect of colours on people’s judgement of the psychological state of those drawing pictures. Hypotheses Hypothesis 1 (.). Drawings by people suffering from emotional distress might be different from those by people without emotional disturbances. Hypothesis 2. If people have to make a ‘forced choice’ between two drawings, they can tell which drawing looks psychologically healthy and which one problematic. Hypothesis 3. Correct answer rate in the chromatic condition will be higher than in the achromatic condition. Part 1: Stimulus production Method Participants A total of 16 pictures were drawn by eight psychologically healthy participants (two male, six female) and six clients (two male, four female) who came to a private counselling centre. The main complaints reported by clients were emotional problems and interpersonal difficulties. None of them had cognitive problems. Two clients drew two pictures (see Fig. 1 3(A)–4(A) and 6(A)–8(A)), whereas the others drew only one. All the participants took the Symptom-Check-List-90-R and the drawing task. They did not participate in the judging task. The Symptom-Check-List-90-R This questionnaire is a 90-question self-report questionnaire that measures psychological symptoms. There are nine major symptom dimensions: somatization, obsessive-compulsive, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoia, and psychoticism. Three additional indices include GSI, PST, and PSDI. The GSI measures current level or depth of symptoms. The PST shows the total number of symptoms. The PSDI is a measure of symptom intensity. Drawing task Drawing materials were white paper measuring 18 × 24 in. and 12 assorted colours of chalk pastels (i.e., red, orange, yellow, light green, dark green, light blue, dark blue, purple, light brown, dark brown, black, and white). Drawing themes were chosen according to a certain set of principles. Although we could not possibly construct mutually exhaustive and exclusive picture stimuli, we used certain considerations for

Fig. 1. Artwork produced by clients (A) and by psychologically healthy participants (B).

J. Jue, S.-M. Kwon / The Arts in Psychotherapy 40 (2013) 115–119 Table 3 Number of correct answers (percentage) and the differences.

Table 1 The drawing themes and the participants in each stimuli production. Pair

Given drawing themes

Clients

Normal participants

1 2

Draw a picture of a tree. Draw an abstract drawing to express your feelings. Draw geometric shapes to express your feelings. Draw a boat on the sea. Draw geometric shapes to express your feelings. Draw how you feel. Draw shapes to express your feelings. Draw lines to express your feelings.

A B

a b

C

c

C D

d e

E F

f g

E

h

3 4 5 6 7 8

117

Pair 1 Pair 2 Pair 3 Pair 4 Pair 5 Pair 6 Pair 7 Pair 8 Total * ** ***

more thorough testing conditions. First, different pictures drawn by the same person must be used. Second, the same theme might be assigned to different participants. Third, a traditional drawing theme, such as tree, must be included. Fourth, a common drawing theme, such as a boat on the sea, could be included. (Note that drawing the sea is a common theme in our native country, which is a peninsula.) Fifth, graphic components of abstract pictures include lines and shapes. Drawing themes for each component separately and integrative abstract themes will be provided. Sixth, we may use the phrase “to express your feelings” when we ask people to draw abstract pictures. We do not use this phrase for concrete drawing themes, because it would be redundant and/or unnecessary. If concrete pictures that are drawn without this directing phrase show significant differences, then the results will offer much more powerful support for our hypothesis. Table 1 shows the given drawing themes and the participants of each pair. All of the participants drew pictures individually, as directed. No instruction was given to consider colours when drawing.

Achromatic condition (N = 112)

Chromatic condition (N = 112)

2

18 (11.6) 47 (41.9) 42 (37.1) 30 (26.7) 78 (69.6) 75 (66.9) 67 (59.8) 31 (27.6) 388 (43.3)

76 (67.8) 85 (75.8) 60 (53.5) 73 (65.1) 92 (82.1) 95 (84.8) 85 (75.8) 51 (45.5) 617 (68.8)

61.66*** 26.63*** 5.83* 33.23*** 4.78* 9.76** 6.63* 7.69** 118.81***

p < 0.05. p < 0.01. p < 0.001.

chromatic condition presented pictures in colour, and the achromatic condition presented the same pictures in monochrome. Picture stimuli were presented in pairs. A total of eight pairs were presented in either a chromatic or an achromatic condition. For instance, if one subject saw pair #1, 2, 4, 5 in the achromatic condition, then s/he saw the rest of the pairs presented in the chromatic condition. All the stimuli were presented only once to each subject. Therefore, the number of answer in each condition per pair is 112. The position of each drawing (left-right) and order of presentation were randomized. Process The instructions were as follows: ‘You will see two pictures at a time. One picture was drawn by a psychologically healthy person, and the other was drawn by a person with diverse psychological problems, including depression, interpersonal sensitivity, and/or hostility. Choose one picture which seems to be drawn by a person with psychological problems’. After subjects chose a picture, the next pair was presented. At the end of the experiment, subjects were asked to explain the reason for their choices.

Results Table 2 shows the means of SCL-90-R indices. The healthy participants had no significantly raised scores in clinical rating scales. On the other hand, clients with diverse psychological symptoms showed much higher levels of depression, interpersonal sensitivity, hostility and/or other clinical symptoms. Fig. 1 shows the drawings of participants.

Results The results are shown in Table 3. The 2 test was used to examine the differences in correct answer rates between the two conditions. All eight pairs showed significant differences in correct answer rates between the two conditions. A total correct answer rate in an achromatic condition was made on a chance level, a half-corrected rate. With colour information, the correct answer rate increased to 68.8%. The most significant difference between the two conditions was discovered in Pair-1, 2 ((1, N = 224) = 61.66, p < 0.001. The subjects who saw Pair-1 in the achromatic condition said that a tree stump and slender tree trunks in picture 1(B) seemed to be the work of an unhealthy person. Others in the chromatic condition said that the blue-green colour use was relaxing compared with the work presented in 1(A), whose colours were intense and provocative. Pair-2 also produced a statistically significant Chi-square score, 2 (1, N = 224) = 26.63, p < 0.001. When the drawings were

Part 2: Judgement Method Subject Subjects in the judging task were 224 individuals aged from 20 to 65. There were 110 males and 114 females. None of them had colour blindness or colour weakness. Stimulus presentation After the drawings were completed, pictures were taken by a digital camera, and presented in pairs on a computer. The Table 2 Means of SCL-90R indices.

Clients Healthy

SOM

OC

IS

DEP

ANX

HOS

PHOB

PAR

PSY

GSI

PST

PSDI

64.62 39.87

64.12 38.62

67.50 41.75

68.62 40.87

65.12 40.37

66.25 39.5

58.00 41.12

67.87 39.12

63.25 41.37

46.00 15.75

65.00 37.25

63.62 37.62

SOM: somatization, OC: obsessive-compulsive, IS: interpersonal sensitivity, DEP: depression, ANX: anxiety, HOS: hostility, PHOB: phobia, PAR: paranoid, PSY: psychoticism, GSI: Global Severity Index, PST: Positive Symptom Index, PSDI: Positive Symptom Distress Index.

118

J. Jue, S.-M. Kwon / The Arts in Psychotherapy 40 (2013) 115–119

Table 4 The characteristics of colour in the drawings. Clients’ artworks

Healthy participants’ artworks

Main colour

6 Red, orange, dark green, light brown, dark brown, black Red

7 Yellow, light green, dark green, light blue, dark blue, purple, black light blue

Number of colours List of colours

3 Red, light brown, black

Main colour

Red, black

7 Red, orange, yellow, light green, dark blue, purple, black evenly used

Pair-3

Number of colours List of colours Main colour

3 Red, yellow, black Black, red

4 Red, yellow, purple, black Red, yellow

Pair-4

Number of colours List of colours

2 Dark blue, black

Main colour

Dark blue

4 Orange, light green, dark green, dark blue Dark blue

Number of colours List of colours

4 Red, dark blue, purple, black

Main colour

Evenly used

Number of colours List of colours

8 Red, orange, yellow, light green, dark green, dark blue, purple, black

Main colour

12 Red, orange, yellow, light green, dark green, light blue, dark blue, purple, light brown, dark brown, black, white Black

Number of colours List of colours

2 Black, white

Main colour

Black

11 Red, orange, yellow, light green, dark green, light blue, dark blue, purple, light brown, dark brown, black Evenly used

Number of colours List of colours

8 Yellow, dark green, light blue, dark blue, purple, light brown, dark brown, black Black

Pair-1

Pair-2

Pair-5

Pair-6

Pair-7

Pair-8

Number of colours List of colours

Main colour

presented in the achromatic condition, 2(B) was judged to be the work of an unhealthy person owing to its small dots. In the chromatic presentation, subjects chose 2(A) as the unhealthy picture, and explained that the intensity of colours and a red-black combination had influenced their decision. The result of Pair-3 showed a statistically significant Chi-square score, though correct answer rates in both conditions were relatively low, 2 (1, N = 224) = 5.83, p < 0.05. Post-experiment inquiry revealed that plaid in 3(B) attracted a lot of attention and affected judges’ choices. The correct answer rates in Pair-4 changed significantly between the two presentation conditions, 2 (1, N = 224) = 33.23, p < 0.001. People who made the wrong choice in the achromatic condition explained that the distinctiveness of boundaries influenced their decisions. Pair-5 and Pair-6 showed a correct answer rate of over 80% in the chromatic condition. The rates were over 65% even in the achromatic condition. Pair-7 also evinced a significant difference in correct answer rates between the two conditions. Although Pair-8 had a relatively low correct answer rate in both conditions, the difference between the two conditions was statistically significant, 2 (1, N = 224) = 7.69, p < 0.01. The subjects in both conditions said that the wedge shape in 8(B) seemed to indicate nervousness and was tension-provoking. Table 4 shows the characteristics of colours in Pairs 1 to 8. These drawings were analyzed according to colour indices: number of colours, list of colours, and main colour. Idiosyncratic colour index was omitted because it is inapplicable to abstract drawings. The colour indices showed the following results: Six out of eight pairs of clients’ drawings had less colour usage than healthy participants’

6 Red, orange, yellow, dark green, dark blue, black Evenly used

Evenly used

5 Red, yellow, dark blue, purple, black

Black, red

drawings. Red and black were frequently used as a main colour in clients’ artworks, whereas healthy drawing participants used colour evenly. Discussion As expected, judging emotional states through comparison of drawings was more accurate in the chromatic condition than in the achromatic. All eight pairs showed statistically significant increases in correct answer rate when two drawings were presented in colour. It is important to note that the achromatic condition showed pictures in monochrome without changing lines, shapes, sizes, spaces and other graphic ingredients. Therefore any increase in the correct answer rate between the two conditions was caused by colour. The results of this study demonstrate that colour conveys information about emotions, and that even lay persons can use the information implicitly or explicitly. Some of stimuli were easy to assess psychological states even in the achromatic condition. Pair-5 and Pair-6 showed 69% and 66% correct answer rates in the achromatic condition respectively. Though the drawings in the achromatic condition eliminated colour information, there was still other information including line, shape, texture, shading, pressure, space, and theme. Unsurprisingly, subjects could tell what kind of feelings and emotions the monochrome pictures expressed. The most interesting result in this study was a huge change in correct answer rates between the two conditions. Pairs 1, 2, 3 and 4 had correct answer rates below 50% in the achromatic condition, and over 50% in the chromatic condition. Since all the

J. Jue, S.-M. Kwon / The Arts in Psychotherapy 40 (2013) 115–119

graphic indicators except colour were the same in both conditions, colour was responsible for these results. In other words, colour is not an auxiliary information source about expressed emotions, but a major determinant in assessing emotions. Therefore, the findings from this study support the third hypothesis, namely colour involves emotional information, which then helps people to estimate emotions. The characteristics of colour usage showed that clients used less colours in their drawings than healthy participants. The list of colours and the main colour in clients’ drawings often included red and/or black. These results are partly in agreement with previous findings, which indicated that less use of colour might reflect depressive states (Wadeson, 1980). The post-inquiry qualitative findings indicate that subjects based their decisions on many aspects of the drawings, including colour. They reported that they felt red and black colour usage was aggressive and anxious. Some subjects insisted that abundant use of black on drawings darkened the images and impressions of drawings. These comments are consistent with the findings that red typically symbolizes anger (Furth, 1988; Malchiodi, 1998), and that even children know that red expresses anger (Zambelli, Clark, & Heegaard, 1989). Red also symbolizes aggression when combined with black (Heller, 2002). A few subjects said that they had made decisions ‘without any reason’ or ‘it just looks like that’. These comments might reflect the implicit effect of colours. According to Davidoff (1991), the reason for the connection of colour with emotion could be innate or a learned reaction. Although every individual has different colour preferences and colour connotations, there is a common use of colour to some degree. In fact, emotional responses to colours have common aspects across cultures, presumably based on natural colours, such as colours of leaves, sky, and darkness. Since the impression and meaning of colours might have universal aspects, lay persons, not only art therapists and art professionals, can appreciate the messages in artwork. Drawings contain inner experiences, which cannot be expressed verbally, and the most important experiences are feelings and emotions. If there is no universality in colour usage, then inner experiences expressed in drawings could not be appreciated except by the artist. Colour has universal aspects, however, and it conveys emotional content. Therefore pictures could constitute communicational channels and have diagnostic value. There were several limitations to the study. Although drawings were made by healthy participants and clients, the stimulus was limited to specific drawings. Future research might include more drawings by other populations. Second, no painters in this study

119

were professional, and their artistry differed. Methodologically, it would be preferable to control for artistic proficiency. References Amos, S. (1982). The diagnostic, prognostic, and therapeutic implications of schizophrenic art. The Arts in Psychotherapy, 9, 131–143. Betts, D. (2003). Developing projective drawing test: Experiences with the Face Stimulus Assessment. Art therapy: Journal of the American Art Therapy Association, 20(2), 77–82. Birren, F. (1988). The symbolism of colour. Secaucus, NJ: Citadel Press. Cohen, B. (1986). The Diagnostic Drawing Series rating guide. Alexandria, VA: Barry Cohen. Cohen, B., Hammer, J., & Singer, S. (1988). Diagnostic Drawing Series: A systematic approach to art therapy evaluation and research. The Arts in Psychotherapy, 15, 11–21. Cohen, B., Mills, S., & Kijak, A. K. (1994). An introduction to the Diagnostic Drawing Series: A standardized tool for diagnostic and clinical use. Art Therapy, 11(2), 105–110. Davidoff, J. (1991). Cognition through colour. Cambridge, MA: MIT Press. Furth, G. (1988). The secret world of drawings. Boston, MA: Sigo Press. Gantt, L. M. (2001). The Formal Elements Art Therapy Scale: A measurement system for global variables in art. Art therapy: Journal of the American Art Therapy Association, 18(1), 51–55. Gantt, L., & Tabone, C. (1998). Formal Elements Art Therapy Scale. Morgantown, WV: Gargoyle Press. Gulbro-Leavitt, C., & Schimmel, B. (1991). Assessing depression in children and adolescents using the Diagnostic Drawing Series modified for children (DDS-C). The Arts in Psychotherapy, 18, 353–356. Hacking, S. (1999). The psychopathology of everyday art: A quantitative study [Doctoral Dissertation]. University of Keele, Sheffield, UK. Available from http://www.musictherapyworld.de/modules/archive/stuff/papers/Hacking.pdf Heller, E. (2002). Wie Farben auf Gefuhl und Verstand wirken. Seoul: Yedam. Jalil, N., Yunus, R. M., & Said, N. S. (2012). Environmental colour impact upon human behavior: A review. Procedia: Social and Behavioral Sciences, 35, 54–62. Lande, R. G., Howie, P., & Chang, A. S. (1997). The art of crime. American Journal of Art Therapy, 36(1), 2–10. Malchiodi, C. (1997). Breaking the silence: Art therapy with children from violent homes (2nd rev.). New York, NY: Brunner/Mazel. Malchiodi, C. (1998). Understanding children’s drawings. New York, NY: Guilford Press. Neale, E. L. (1994). The children’s diagnostic drawing series. Art Therapy: Journal of American Art Therapy Association, 11(2), 119–126. Oster, G., & Gould, P. (1987). Using drawings in assessment and therapy: A guide for mental health professionals. Bristol, PA: Brunner/Mazel. Pianetti, C., Palacios, M., & Elliott, L. (1964). Significance of colour in the drawings of chronic schizophrenics. American Journal of Occupational Therapy, 18, 137–140. Suenaga, T. (1998). Colour psychology. Seoul: Yekyong. Ulman, E. (1992). A new use of art in psychiatric diagnosis. American Journal of Art Therapy, 30, 78–88. Wadeson, H. (1971). Characteristics of art expression in depression. Journal of Nervous and Mental Disease, 153(3), 197–204. Wadeson, H. (1980). Art psychotherapy. New York, NY: Wiley. Yokota, M., Ito, N., & Shimizu, O. (1999). A study of coloured tree drawings of patients with schizophrenia: Part 2. Seishin Igaku Clinical Psychiatry, 41(5), 469–476. Zambelli, G. C., Clark, E. J., & Heegaard, M. (1989). Art therapy for bereaved children. In H. Wadeson, J. Durkin, & D. Perach (Eds.), Advances in art therapy (pp. 60–80). New York, NY: John Wiley & Sons.