Emotion 2009, Vol. 9, No. 5, 734 –738
© 2009 American Psychological Association 1528-3542/09/$12.00 DOI: 10.1037/a0016818
Context Specificity of Implicit Preferences: The Case of Human Preference for Red Markus A. Maier
Petra Barchfeld
Stony Brook University
University of Munich
Andrew J. Elliot
Reinhard Pekrun
University of Rochester
University of Munich
Three experiments were conducted on color preference using a spontaneous selection paradigm with infant participants. Experiment 1 demonstrated that participants prefer red over green in a friendly laboratory environment. Experiment 2 demonstrated that participants’ preference for red varies with the context in which the color is presented: Red is preferred in a hospitable context (following a happy face), but not in a hostile context (following an angry face). The opposite pattern was found for the control color green. Experiment 3 used the same context manipulation, but a second control color, gray, was added to clearly examine whether context affects preference for red only. As predicted, given a second alternative choice, context-dependent preference for red, but not green or gray, was found. These results represent the first evidence of context moderation in the color preference literature. Keywords: color preference, red, context, avoidance, approach
Castain, 2004). This preference for red appears to be age dependent, however, in that it is no longer present by the time children enter first grade, and is not seen again during the remainder of the life course (e.g., Norman & Scott, 1952).
Research on color preference has a long history in scientific psychology, dating back to work by Jonas Cohn (1894) in Wilhelm Wundt’s laboratory. To date, research in this area has focused exclusively on general color preferences, implicitly assuming that color preference is invariant across contexts. In the present work, we drew on recent research on the color red conducted with both human participants (Elliot, Maier, Moller, Friedman, & Meinhardt, 2007; Elliot & Niesta, 2008) and nonhuman animals (Dominy & Lucas, 2001; Setchell & Wickings, 2005) to argue for and illustrate the importance of attending to context in the study of color preference.
Color in Context In all color preference research, whether conducted with infants, older children, or adults, the issue of context has been widely ignored.1 In the present work, we considered a basic feature of context that plays an integral role in the functioning of all animate organisms—that of motivational valence (Schneirla, 1959). Research clearly indicates that all encountered stimuli are categorized at a rudimentary level as positive (hospitable) or negative (hostile), and that these categorizations establish a motivational context that influences affect, cognition, and behavior (Cacioppo & Gardner, 1999; Lang, 1995). For example, when participants are primed with a happy face, they evaluate novel objects (e.g., Chinese idiographs) more positively, whereas when they are primed with an angry face, they evaluate such objects less positively (Murphy & Zajonc, 1993). We posit that on the basis of empirical work on signal coloration in nonhuman animals and, to a lesser degree, on research on color and psychological functioning in humans, context influences color preference, particularly evaluation of the color red, in similar fashion. Research on a variety of nonhuman primates indicates that in appetitive (hospitable) contexts, red has a positive connotation and
Research on Color Preference Research on color preference is as voluminous as it is aged (Whitfield & Wiltshire, 1990), and has been conducted with individuals of all age groups. Some studies have been conducted with infants younger than 2 years of age. These studies show that infants have a preference for red over other chromatic colors; some studies yield strong evidence for this red preference (e.g., Adams, 1987; Bornstein, 1975), whereas others provide weak or partial evidence (e.g., Garth & Porter, 1934; Teller, Civan, & Bronson-
Markus A. Maier, Department of Psychology, Stony Brook University; Petra Barchfeld and Reinhard Pekrun, Department of Psychology, University of Munich; Andrew J. Elliot, Department of Clinical and Social Sciences in Psychology, University of Rochester. Thanks are extended to the following individuals who helped in the data collection process: Phoebe von Seyerl and Martina Gerl. Correspondence concerning this article should be addressed to Markus A. Maier, Department of Psychology, Stony Brook University, Stony Brook, NY 11794-2500. E-mail:
[email protected]
1
By context, we mean psychological context. Some research has been done investigating the influence of the physical context of color presentation in predicting color preference (e.g., Bornstein, Kessen, & Weisskopf, 1976). 734
BRIEF REPORTS
facilitates approach-oriented behavior. For example, in mating contexts, red on the hindquarters of a female signals sexual availability, and facilitates mounting behavior in male conspecifics (Nunn, 1999). Similarily, in nutrition contexts, red leaves activate grasping behavior, presumably because of the higher amount of protein they carry (Dominy & Lucas, 2001). On the other hand, research on these same animals indicates that in aversive (hostile) contexts, red has a negative connotation and reduces approachoriented behavior. For example, in competitive contexts, red on the chest of a male signals dominance and attack readiness, and keeps lower status conspecifics from approaching (Setchell & Wickings, 2005). Thus, in many nonhuman animals, the hospitable or hostile nature of the context at hand dictates whether red is a desirable or undesirable signal. It is interesting that recent experiments with human participants have yielded conceptually related findings. Human males show signs of being appetitively motivated on viewing red on females in mating contexts, much like nonhuman male primates (Elliot & Niesta, 2008). In addition, human males and females show signs of being aversively motivated on viewing red in competitive contexts, much like nonhuman male and female primates (Elliot, Maier, Binser, Friedman, & Pekrun, 2009; Elliot et al., 2007). In sum, there are hints that red carries different meanings in different contexts for both humans and nonhuman animals. However, context specificity has yet to be examined with regard to the issue of color preference. In the studies presented herein, we addressed this oversight by manipulating context and examining its impact on preference for the color red. In doing so, we concentrated on the color preference of infants. As noted earlier, the existing research indicates that the affective meaning of red changes with age, being positive before elementary school age and becoming less positive thereafter. Thus, by using an age group for which preference for red can be expected as the default value, we put our hypothesis regarding contextspecific variation of preference for red to the most stringent test possible. Our hypothesis was that a context manipulation would influence infants’ preference for red. Specifically, we predicted that in the presence of a happy (hospitable) face red would be preferred because it signals a potentially desirable outcome, paralleling the appetitive context in prior research, but that in the presence of an angry (hostile) face, red would not be preferred because it signals a potentially undesirable outcome, paralleling the threatening context in prior research.
Experiment 1 The aim of this study was to conceptually replicate previous findings indicating a preference for red in infancy in a standard laboratory context. We examined infant color preference using a spontaneous selection paradigm, with the grasping of one of two colored objects placed in front of the infant as the preference measure. Most existing studies on infant color preference have used looking time or frequency as the dependent measure. Although infants may look at a color because they find it pleasant, they may also look at it for other reasons, such as curiosity or wariness (Adams, 1987). Spontaneous selection of a colored object seems a more direct indicator of preference in that it clearly represents an approach behavioral response; recent research has
735
shown that choosing an object is indeed indicative of preference (Savani, Markus, & Conner, 2008). We used spontaneous selection as our dependent measure in the present research. We examined preference for red relative to green, a chromatic color considered the opposite of red in well-established color models (Kaiser & Boynton, 1996). Green has been used successfully as a control color in several previous studies (see Elliot & Maier, 2007). In accord with the tendency in the looking time and frequency data, we anticipated that infants would exhibit a preference for red over green by selecting it more often in a spontaneous selection paradigm in the friendly laboratory environment typically used in infant research.
Method Participants. Twenty-two infants (12 girls; mean age ⫽ 12.71 months, SD ⫽ 2.48) successfully completed the study. All infants were members of a day care center. Materials. Two objects made of Lego bricks were used as the choice stimuli. The objects were 15 cm in height and 10 cm in width, and were a unique shape (similar to an asymmetric number 8) that bore no resemblance to any object commonly encountered by the participants. One object was created from red Lego bricks, whereas the other was created from green Lego bricks. A Minolta CS-100 spectrophotometer was used to determine the parameters of the colors from the spectral data: red (LCh [45.8; 77.9; 38.45], CIE-Yxy [15.1; 0.60; 0.34]), and green (LCh [45.6; 57.0; 148.33], CIE-Yxy [15.0; 0.27; 0.52]). As can be seen from these values, chroma (C) was different for the two colors, but lightness (L) was essentially identical (Stokes, Fairchild, & Berns, 1992). Procedure. The experiment was conducted in a quiet day care center room highly familiar to infants. All chromatic color stimuli were removed from the room prior to the experimental session. In this and all subsequent experiments, the experimental room was illuminated by cool white fluorescent lighting. Two experimenters and one caregiver from the day care center were present during the experiment. Both experimenters wore blue, brown, or gray clothes. The experimenters and the caregiver were blind to the experimental hypothesis. The caregiver sat at a small table, and the participant sat on the legs of the caregiver directly in front of the table. The caregiver was instructed not to interact with the participant during the session. One experimenter stood in front of the participant on the opposite side of the table, and the other stood to one side of the participant. The table contained the two objects covered by a white cloth. The experimenters began by interacting with the participant in a friendly manner for several minutes to put the child at ease. Then, the experimenter in front of the participant directed the child’s attention to the cloth on the table as the caregiver moved the participant within reach of the cloth (and underlying objects). When the participant looked at the cloth, the experimenter removed it and stated, “Look what I brought along.” The objects had been placed side by side, and the position (left vs. right) of the red and green objects was counterbalanced across participants. The dependent measure was the color of the object that the participant selected first, which was recorded by the experimenter to the side of the child. The experiment was stopped after the initial selection.
BRIEF REPORTS
736 Results and Discussion
A chi-square test of independence was calculated to determine participants’ color preference. The analysis yielded a significant effect, 2(1, N ⫽ 22) ⫽ 4.55, p ⬍ .05. Participants exhibited a preference for the red object (n ⫽ 16) over the green object (n ⫽ 6). No gender or order effects were found ( ps ⬎ .34). The findings indicate that red is preferred over a control color in a friendly context.
Experiment 2 Experiment 2 examined the influence of context on color preference by exposing participants to a black-and-white photo of an happy or angry face immediately prior to selection of the red or green colored object. Prior research, including that using EMG measurement of automatic affective responding (Balaban, 1995), has shown that infants as young as 5 months of age exhibit the expected positive and negative reactions to briefly presented happy and angry faces, respectively. We predicted that infants would be more likely to select the red object in the happy, hospitable context, compared with the angry, hostile context. In addition to the inclusion of context, we attended to a potential alternative explanation for the Experiment 1 findings. Color varies according to three parameters— hue, chroma, and lightness. Given that chroma was not controlled in Experiment 1, the observed results could be due to a difference in this color parameter rather than a difference in hue per se (see Valdez & Mehrabian, 1994). Thus, in addition to lightness, we equated red and green on chroma in this experiment to eliminate this possibility. Only two existing infant studies have addressed this issue with comparable rigor (Adams, 1987; Zemach, Chang, & Teller, 2007), both of which used looking behavior to assess color preference.
Procedure. The procedure was the same as that used in Experiment 1, with the exception that immediately before the objects were revealed, one of the two faces was presented to the infant (a between-subjects manipulation; 12 were shown the happy face, and 15 were shown the angry face). The face was presented by the experimenter across the table from the participant. The picture was initially on the floor, face down next to the experimenter, so that the experimenter could not see the picture type. When the time came for the context manipulation, the experimenter picked up the picture and held it in front of her in full view of the child at a distance of 70 cm. The experimenter waited until the participant looked at the picture and then presented it for an additional 2 s. During this time, the picture was resting in an upright position on the tabletop. The experimenter who presented the picture remained blind to picture type throughout the experiment (she saw the back of the picture only); both experimenters and the caregiver were blind to the experimental hypotheses.
Results and Discussion A chi-square test of independence was calculated to determine whether participants’ color preference varied by context. The analysis yielded a significant effect, 2(1, N ⫽ 27) ⫽ 13.23, p ⬍ .01. Participants in the hospitable context had a clear preference for the red object (n ⫽ 10) compared with the green object (n ⫽ 2), 2(1, N ⫽ 12) ⫽ 5.33, p ⬍ .05, whereas participants in the hostile context had a clear preference for the green object (n ⫽ 13) compared with the red object (n ⫽ 2), 2(1, N ⫽ 15) ⫽ 8.07, p ⬍ .01. No gender or order effects were found ( ps ⬎ .15). The results make clear that preference for the colors red and green changes with the context in which they are presented, with red being more preferred in a hospitable situation and green being more preferred in a hostile situation.
Experiment 3
Method Participants. Twenty-seven infants (9 girls; mean age ⫽ 12.56 months, SD ⫽ 1.51) successfully completed the study. All infants were members of a day care center. Materials. The objects used in this experiment were similar in shape and size to those used in Experiment 1 (18 cm ⫻ 11 cm), but were composed of red and green paperboard rather than red and green Lego bricks. The colors were printed on paperboard to allow the red and green hues to be equated on chroma and lightness. An Eye One Pro spectrophotometer was used to determine the parameters of the colors from the spectral data: red (LCh [49.2; 49.6; 28.0], CIE-Yxy [17.76; 0.49; 0.33]) and green (LCh [49.8; 50.5; 159.4], CIE-Yxy [18.25; 0.25; 0.45]). As can be seen from these values, chroma (C) was essentially identical for the two colors, as was lightness (L). Two black-and-white versions of pictures from the International Affective Picture System (Lang, Bradley, & Cuthbert, 1995) were used for the context manipulation. The pictures consisted of the same man displaying either a happy face (the hospitable context) or an angry face (the hostile context). Lang et al. (1995) provide normative data supporting the affective categorizations of the two pictures used herein. Presenting happy and angry facial expressions to infants to induce affective valence is well established (Balaban, 1995).
In Experiment 3, we added a third colored object to the preference task to investigate the generalizability and replicability of our findings to a different control color. In Experiment 2, a binary choice task was used to measure context-specific preference for red; thus, the number of alternative choices was quite limited. Furthermore, the data from Experiment 2 suggested context specificity for both red and green, but context-specific variation was only predicted for red. However, the context-specific variation for green may simply have been caused by variation in the preference for red because a decrease in preference for red always leads to an increase in preference for green (and vice versa) in a binary choice task. Therefore, in Experiment 3, we provided participants with an additional alternative choice to reduce the interdependency among the color options and, on the basis of our theorizing, we expected context-specific variation to emerge for red only. We used gray as our second control color; like green, gray has been used successfully as a control color in prior research (Maier, Elliot, & Lichtenfeld, 2008).
Method Participants. A total of 40 infants (25 girls) successfully completed the study. The mean age of participants was 17.75 months
BRIEF REPORTS
(SD ⫽ 2.49), with a range of 15 to 26 months. All infants were members of a day care center. Materials. The materials used in the experiment were the same as those used in Experiment 2, with one critical exception: A third, gray, object was used together with the red and green objects. The gray color (LCh [49.4; 2.3; 140.2], CIE-Yxy [17.90; 0.35; 0.37]) was equated on lightness with the red and green colors; gray is an achromatic color, meaning equating on chroma is not relevant. Procedure. The procedure was the same as in Experiment 2, except participants in this experiment had a choice among three colored objects (instead of two): red, green, and gray. The position of the colored objects was counterbalanced across participants.
Results and Discussion A chi-square test of independence was calculated to determine whether participants’ color preference varied by context. The analysis yielded a significant effect, 2(2, N ⫽ 40) ⫽ 8.24, p ⬍ .02. As can be seen in Figure 1, participants in the hospitable context had a clear preference for the red object (n ⫽ 15) compared with participants in the hostile condition (n ⫽ 6), 2(1, N ⫽ 21) ⫽ 3.86, p ⬍ .05. In contrast, neither preference for green, 2 ⬍ 2.78, nor preference for gray, 2 ⬍ 1.60, showed a significant difference as a function of context. No gender or order effects were found ( ps ⬎ .18). In sum, the results from this experiment supported our predictions: The color red was more preferred in a hospitable compared with a hostile context, but context did not affect preference for the colors green or gray.
General Discussion Our results indicate that infants prefer the color red to green when allowed to spontaneously choose between the two in a standard, friendly environment. More provocatively, our findings indicate that infants’ preference for red changes with the context in which it is presented. Specifically, in a hospitable context, red is preferred, whereas in a hostile context, red is not preferred. Context-specific variation in preference was not observed for either green or gray. This represents the first evidence in the infant literature for context-specific color preference, and the first direct experimental evidence in the literature on color preferences more generally. It now seems reasonable to speculate that the preference for red in infancy described in the literature (e.g., Adams, 1987; Bornstein, 1975) might be restricted to experimental situations in
737
which infants feel safe, which is likely the case in the majority of the existing studies. On the other hand, any change in infants’ feelings of safeness (e.g., testing without a familiar other present) may produce a change away from a preference for red, much as we showed herein using an angry face stimulus. Although our findings do not provide direct evidence on the origin of the context-specific variation of preference for red in infancy, we do think it is reasonable to raise the possibility that the observed findings may reflect inherent, rather than learned, response tendencies. Future research with younger infants (e.g., 4 months of age) for whom color vision has just developed would shed further light on this question. The temptation to construe our findings in terms of basic tendencies is reinforced by the parallel between our results and those on signal coloration that served as a primary impetus for our experiments. Our findings indicate that infants, like chimpanzees and other animals, respond favorably to red in a hospitable context (Dominy & Lucas, 2001; Nunn, 1999) and unfavorably to red in an aversive context (Setchell & Wickings, 2005). Furthermore, recent experiments with adult participants have yielded conceptually related findings (Elliot et al., 2007; Elliot & Niesta, 2008). Of course, strong statements about the origins of human response to red are clearly premature. Additional, converging evidence is needed before any definitive con¨ hman & Mineka, clusions about origin issues are warranted (see O 2001, for a conceptual parallel). The lack of attention to context in the color preference literature is surprising to us because other types of preferences, such as those for food, pictures, tasks, and so forth, are clearly recognized as varying from situation to situation (Eagley & Chaiken, 1993; Elliot & Harackiewicz, 1996; Lang, 1995). Indeed, preferences may be defined as evaluative judgments made by individuals in specific psychological situations (Tesser & Martin, 1996), and it seems straightforward to posit that color preferences, like other preferences, depend on the psychological situation in place at a given time. The present work documents one set of situations in which this is the case, and opens the door for follow-up research on other colors and other contexts. On the surface, a color preference may seem to simply represent what is “pleasing to the eye.” Our research highlights the need for a deeper consideration of the nature of color preference in that it suggests that such preferences may be connected to deeper meanings, emotions, and motivations (see also Norman & Scott, 1952). That is, color preference may actually reflect basic approach and avoidance motivations that vary according to the degree of safety or threat in the immediate environment. From this perspective, research on color preference is about much more than aesthetics; it is about the study of important psychological processes using a metric that can serve as a common currency across species and across the human life span.
References
Figure 1. iment 3.
Object preference as a function of color and context in Exper-
Adams, R. J. (1987). An evaluation of color preferences in early infancy. Infant Behavior and Development, 10, 143–150. Balaban, M. T. (1995). Affective influences on startle in five-month-old infants: Reactions to facial expressions of emotion. Child Development, 66, 28 –36. Bornstein, M. H. (1975). Qualities of color vision in infancy. Journal of Experimental Child Psychology, 19, 401– 419. Bornstein, M. H., Kessen, W., & Weiskopf, S. (1976). Color vision and hue
738
BRIEF REPORTS
categorization in young human infants. Journal of Experimental Psychology: Human Perception and Performance, 2, 115–129. Cacioppo, J. T., & Gardner, W. L. (1999). Emotion. Annual Review of Psychology, 50, 191–214. Cohn, J. (1894). Experimentelle Untersuchungen u¨ber die Gefu¨hlsbetonung der Farben, Helligkeiten und ihrer Kombinationen [Experimental investigation of the emotional effect of color, lightness, and their combinations]. Philosophische Studien, 10, 562– 603. Dominy, N. J., & Lucas, P. W. (2001, March 15). Ecological importance of trichromatic vision to primates. Nature, 410, 363–366. Eagley, A., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX: Harcourt Brace Jovanovich. Elliot, A. J., & Harackiewicz, J. M. (1996). Approach and avoidance achievement goals and intrinsic motivation: A mediational analysis. Journal of Personality and Social Psychology, 70, 461– 475. Elliot, A. J., & Maier, M. A. (2007). Color and psychological functioning. Current Directions in Psychological Science, 16, 250 –254. Elliot, A. J., Maier, M. A., Binser, M. J., Friedman, R., & Pekrun, R. (2009). The effect of red on avoidance behavior. Personality and Social Psychology Bulletin, 35, 365–375. Elliot, A. J., Maier, M. A., Moller, A. C., Friedman, R., & Meinhardt, J. (2007). Color and psychological functioning: The effect of red on performance attainment. Journal of Experimental Psychology: General, 136, 154 –168. Elliot, A. J., & Niesta, D. (2008). Romantic red: Red enhances men’s attraction to women. Journal of Personality and Social Psychology, 94, 1150 –1164. Garth, T. R., & Porter, E. P. (1934). The color preferences of 1,032 young children. American Journal of Psychology, 46, 448 – 451. Kaiser, P. K., & Boynton, R. M. (1996). Human color vision (2nd ed.). Washington, DC: Optical Society of America. Lang, P. J. (1995). The emotion probe: Studies of motivation and attention. American Psychologist, 50, 372–385. Lang, P. J., Bradley, M., & Cuthbert, B. (1995). International Affective Picture System (IAPS): Technical manual and affective ratings. Gainesville: NIMH Center for the Study of Emotion and Attention, University of Florida. Maier, M. A., Elliot, A. J., & Lichtenfeld, S. (2008). Mediation of the negative effect of red on intellectual performance. Personality and Social Psychology Bulletin, 34, 1530 –1540. Murphy, S. T., & Zajonc, R. B. (1993). Affect, cognition, and awareness:
Affective priming with optimal and suboptimal stimulus exposures. Journal of Personality and Social Psychology, 64, 723–739. Norman, R., & Scott, W. (1952). “Color and affect”: A review and semantic evaluation. Journal of General Psychology, 46, 185–223. Nunn, C. L. (1999). The evolution of exaggerated sexual swellings in primates and the graded-signal hypothesis. Animal Behaviour, 58, 229 – 246. ¨ hman, A., & Mineka, S. (2001). Fears, phobias, and preparedness: O Toward an evolved module of fear and fear learning. Psychological Review, 108, 483–522. Savani, K., Markus, H. R., & Conner, A. L. (2008). Let your preference be your guide? Preferences and choices are more tightly linked for North Americans than for Indians. Journal of Personality and Social Psychology, 95, 861– 876. Schneirla, T. (1959). An evolutionary and developmental theory of biphasic processes underlying approach and withdrawal. In M. Jones (Ed.), Nebraska Symposium on Motivation (pp. 27–58). Lincoln: University of Nebraska Press. Setchell, J. M., & Wickings, E. J. (2005). Dominance, status signals, and coloration in male mandrills (Mandrillus sphinx). Ethology, 111, 25–50. Stokes, M., Fairchild, M. D., & Berns, R. S. (1992). Precision requirements for digital color reproduction. ACM Transactions on Graphics, 11, 406 – 422. Teller, D. Y., Civan, A., & Bronson-Castain, K. (2004). Infants’ spontaneous color preferences are not due to adult-like brightness variations. Visual Neuroscience, 21, 397– 401. Tesser, A., & Martin, L. (1996). The psychology of evaluation. In E. T. Higgins & A. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 400 – 432). New York: Guilford Press. Valdez, P., & Mehrabian, A. (1994). Effects of color on emotions. Journal of Experimental Psychology: General, 123, 394 – 409. Whitfield, T. W., & Wiltshire, T. J. (1990). Color psychology: A critical review. Genetic, Social, and General Psychology Monographs, 116, 387– 412. Zemach, I., Chang, S., & Teller, D. Y. (2007). Infant color vision: Prediction of infants’ spontaneous color preferences. Vision Research, 47, 1362–1367.
Received November 11, 2008 Revision received April 9, 2009 Accepted May 18, 2009 䡲
