This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
The research exploited a unique architectural setting of a university residence hall composed by six separate buildings that matched for every architectural detail and differed only for the interior color (violet, blue, green, yellow, orange, and red). Four hundred and forty-three students living in the six buildings for an average of 13.33 months participated in a study that assessed color preference (hue and lightness), lightness preference, and the effects of color on studying and mood. The results showed a preference for blue interiors, followed by green, violet, orange, yellow, and red. A preference bias was found for the specific color in which the student lived. Gender differences emerged for the preference of blue and violet. Room-lightness was significantly affected by the interior color. Room ceiling was preferred white. Blue as interior color was considered to facilitate studying activity. The use of differentiated colors in the six buildings was evaluated to significantly facilitate orienting and wayfinding. A significant relation was found between a calm mood and preference for blue.
Keywords: color, chromatic preference, lightness preference, architecture, interior designAlthough color is a ubiquitous property of every architectural surface, evidence-based research on chromatic preference in architecture and psychological effects of color as a function of the architectural design of a space is still sparse. This study exploited a unique architectural setting of a university residence hall for long-term student accommodation, composed by six separate buildings that matched for every design feature with the only exception of interior color. Each building interior was characterized by a specific color for walls, ceiling, and floor in both common spaces and students’ rooms. The colors were: violet, blue, green, yellow, orange, and red. Students were independently assigned to the different colors by the residence administration. Therefore, this architectural setting resulted as an in vivo experiment that allowed a controlled assessment of students’ color preferences, their satisfaction with the color and lightness level of the building they lived in, and their assessment of the building color effect on their mood, studying activity, orienting within the residence.
Colors have three basic perceptual attributes: hue, saturation, and lightness (Hunt and Pointer, 2011). Hue is the phenomenological correspondent of wavelength within the visible-light spectrum. Saturation (chroma) describes the intensity or purity of a hue, whereas lightness (value) varies according to the relative presence of black or white in the color. At the lower extreme of saturation lie the achromatic colors gray, black, and white. Many models exist that map colors along these attributes in two-dimensional or three-dimensional spaces. Some of these models are perceptually uniform, and match human-color perception (e.g., Munsell, CIE Lab), whereas others are not perceptually uniform, and were developed to map colors for specific technical domains (e.g., RGB, HSV, HSL, HSB, CMYK).
Color preferences were mainly investigated manipulating hue, starting from the pioneering work by Eysenck (1941) who established a universal preference hierarchy in colors. According to his study the most preferred color was blue, followed by red, green, violet, orange, and yellow. This finding agreed with those obtained by Granger (1952) and Guilford and Smith (1959) who found the highest preference ratings for the blue-green hues and the lowest for yellow and yellow-green hues. These results were further confirmed by Granger (1955), Dittmar (2001), Bakker et al. (2013), and Schloss et al. (2013). Hue preferences in adults follows a relatively smooth curvilinear function in which cool colors (green, cyan, blue) are generally preferred to warm colors (red, orange, yellow) (Palmer et al., 2013). Focusing on color saturation, Palmer et al. (2013), in a review on color preference studies, concluded that, in general, more colorful and saturated colors are preferred to less vivid color. Saturation interacts with preferences for lightness so that yellow is preferred at high lightness levels, red and green at medium lightness levels, and blue and purple at low lightness levels (Guilford and Smith, 1959). Dark shades of orange (browns) and yellow (olives) tend to be strongly disliked relative to lighter, equally saturated oranges and yellow (Guilford and Smith, 1959; Palmer and Schloss, 2010). Color preference in these studies was assessed rating preselected color patches (either as physical colored chips, or presented on computer monitor), or asking participants to imagine colors, and was not referred to specific objects.
The extent to which these global and abstract color preferences could be applied to specific contexts was the focus of different studies. For example, Taft (1997) compared the abstract semantic ratings of color samples with those of the same colors applied to a variety of familiar objects (e.g., sofa, modern chair, antique chair, bicycle, cheese slicer, and computer), finding a good correspondence between the two sets of ratings. Overall, he found that only in the 4% of cases the color on the sample was judged different for attractiveness from the same color on an object. The specificity of color preference for specific objects was explained in terms of appropriateness of the color-object association based on people experience. Some objects, in fact, can be found in a wide variety of colors (e.g., bicycles), whereas many objects appear in a very limited range of colors (e.g., computers, smartphones). Schloss et al. (2013) showed eight hues, each at two levels of saturation and two levels of lightness, in addition to five achromatic colors (black, white, and three shades of gray). Participants had to rate the preference of each color contextless on simple patches, and with reference to different objects, both imagined and depicted (e.g., car, t-shirt, walls, sofa etc.). The results showed that people preferred more saturated colors when evaluating simple patches than real objects. They also preferred darker colors for objects (e.g., t-shirts, scarfs, and couches) compared to participants’ general preferences, with the exception of walls that were preferred with lighter colors. Furthermore, wall colors were preferred lighter in the imagined condition compared to the depicted condition. Jonauskaite et al. (2016) investigated context-specific color preferences comparing abstract color preferences, imagined interior walls, and imagined t-shirts. They used an unrestricted color selection approach with three-color dimensions (i.e., hue, chroma and lightness). Abstract colors were preferred with more chroma, whereas lighter colors were preferred for walls, and darker colors were preferred for t-shirts.
In the specific architectural context, Kunishima and Yanase (1985) investigated the visual effects of wall colors in living rooms. Architectural students had to evaluate living room models differing in color. A factor analysis highlighted three main dimensions: “activity,” “evaluation,” and “warmness.” “Activity” was mostly affected by the brightness of the wall color, “evaluation” by the saturation, and “warmness” by the hue.
The impact of light and color on psychological mood in work environments was investigated by Küller et al. (2006) in a large-scale study that involved 988 persons from different countries. The presence of some colors, in comparison to a no-color, or neutral-color condition, resulted in a more positive worker’s mood. The use of very saturated colors, to the contrary, had a negative effect on mood.
Several studies investigated the role of sex and culture to test the universality of color preference (Choungourian, 1968; Saito, 1994, 1996; Ou et al., 2004, 2012; Hurlbert and Ling, 2007; Al-Rasheed, 2015). A study on sex differences found a peak for the blue-green in the preference pattern of males and a peak for the reddish-purple region for females but when Chinese and British participants were analyzed separately the sex differences emerged only in the British subpopulation (Hurlbert and Ling, 2007).
Taylor et al. (2013a) pointed out that previous studies focused mainly on industrialized cultures, and they decided to compare color preferences of British adults to those of Himba adults, who belong to a non-industrialized culture in rural Namibia. Results suggested that predictive models proposed in previous studies cannot account for the differences observed in the two populations.
Another cross-cultural study found significant differences between a population from Poland and a population from Papua (Sorokowski et al., 2014), even if sex patterns had a much higher effect size than cultural difference. In fact, although preferences observed in the two populations were different, the differences observed in the preference patterns of males and females were comparable in the two samples.
Some studies also investigated the relation between color preferences and age (Teller et al., 2004; Zemach et al., 2007; Franklin et al., 2008, 2010; Taylor et al., 2013b). For hue preference there is good agreement between different age categories. In particular, both infants and adults tend to show a preference for blue and a dislike for greenish-yellow (Teller et al., 2004; Zemach et al., 2007; Franklin et al., 2008, 2010). Palmer and Schloss (2010) found significant differences for lightness and saturation (infants tend to prefer saturated and light hues). In comparative studies, the preference for blue was also confirmed in rhesus monkeys (Humphrey, 1972; Sahgal et al., 1975), and pigeons (Sahgal and Iversen, 1975).
Different accounts have been proposed to explain color preference. According to Hurlbert and Ling (2007) color preference is rooted in the cone-opponent contrast neural mechanisms which encode colors. Human color vision is in fact based on two cone-opponent systems, loosely called “red-green” and “blue-yellow.” The red-green system responds to the difference between long-wavelength-sensitive cone responses (L) and middle-wavelength-sensitive (M) responses (L–M), while the blue-yellow system differences short-wavelength-sensitive (S) cones with a combination of L and M cones [S – (L + M)]. The blue-yellow system accounts for the greatest variance (44.5%) for color preference across the population, with blue hues that are preferred over yellow hues. To the contrary, the red-green system accounts mainly for sex differences, with females that prefer colors with “reddish” contrast against the background in comparison to males (Hurlbert and Ling, 2007).
In another perspective, color preference could be grounded on emotional associations of colors. Colors are strictly associated to specific emotional states (Ou et al., 2004), and if an emotional state is perceived as pleasant then indirectly the pleasantness is transferred to the color. According to this theory, active, light, and cool colors are being preferred over passive, heavy, and warm ones. This theory, however, fails to explain why although blue is associated with sadness it is the most preferred color, and why yellow which is associated with joy, is less preferred than blue.
According to the ecological valence theory (EVT, Palmer and Schloss, 2010) color preferences arise from people’s average affective responses to color-associated objects, so that people like colors strongly associated with objects they like and dislike colors strongly associated with objects they dislike. For example, since water is important for surviving and water tends to be blue, blue is largely appreciated; similarly, since rotten food is dangerous for our health and rotten food tend to be greenish-yellow, this color is largely unappreciated. The EVT is able to explain both the universal trends and the minor variations: blue is probably appreciated in every culture while red is generally less appreciated, but for example the lucky effect that Chinese culture associate to this color make it more appreciable in China compared to other countries. The authors of the EVT estimated that the affective valence association was able to account for the 80% of variance in color preference ratings over 32 different colors.
Few controlled studies have investigated psychological and physiological effects of specific color exposure. For example, Jacobs and Hustmyer (1974) measured the physiological activation during a 1-min exposure to four different colors. Considering the galvanic skin response, red was significantly more arousing than other colors. Küller et al. (2009) compared psychological and physiological effects of a gray, red, and blue room. The results showed that the red room increased the brain arousal level (assessed as percentage of alpha waves). This effect was particularly significant in introvert persons or persons that were in a negative mood. Red was also found to be associated with a higher probability of winning a sport competition (Hill and Barton, 2005), to performance impairment on achievement tasks due to avoidance motivation (Elliot et al., 2007; Mehta and Zhu, 2009), and to performance enhancement on detail-oriented tasks (Mehta and Zhu, 2009).
Kwallek et al. (1996) compared nine monochromatic office interior colors in a between-subjects study in which university students performed a proofreading task in one office for a total permanence of 45 min. The nine office colors varied for two levels of saturation (high/low), and two levels of lightness (dark/light). Pre and post mood change and color preferences were also recorded. The proofreading task performance was not affected by office color, whereas errors were higher in the white office in comparison to the blue and red offices, even if it cannot be excluded that this difference could stem from cognitive differences between the groups in the different conditions. Higher saturated color offices resulted in higher vigor scores for mood. Lightness and coolness or warmth of the office color did not influence mood. Pleasantness for the office color differed significantly between the groups. Individuals preferred to work in beige and white rooms than in orange and purple offices. In terms of whether they liked the office color, individuals in the green and red offices preferred their office color more than individuals in the yellow and orange offices. Participants in the white, beige, blue, and gray offices liked the color of their offices more than participants in the orange office. Concerning the distracting effect of the color, participants in the purple, orange, red, yellow office colors reported that their colors were more distracting compared to participants in the green, gray, beige, and white offices. Purple and yellow office colors were rated as the most distracting, and white as the least distracting.
In the context of criminal detention holding cells Pellegrini et al. (1981) found no difference in the incidence of aggressive officer-arrestee encounters after changing the cell color from pale blue to hot pink.
Independently from the influence of color on behavior, people strongly tend to associate colors to specific semantic clusters (Sutton and Altarriba, 2016). Bright colors (e.g., white, pink) are often associated to positive emotions whereas dark colors (e.g., black, brown) tend to be associated with negative emotions (Hemphill, 1996). Furthermore, we tend to infer the valence of a stimulus on the basis of brightness (Meier et al., 2004). Individuals were faster to categorize positive words when they appeared in white than when they appeared in black, with an opposite trend for negative words. Color associations are often cross-modal (Spence, 2011), and the most important cross-modal association is the distinction between cold and warm colors (Ho et al., 2014).
Most of the literature that we have so far reviewed defined color effects and preferences exposing participants to colors via computer screens or using colored patches, or asking participants to imagine specific colors; furthermore, the exposure time to colored settings was in general very short (Elliot and Maier, 2014). The more realistic setting in an architectural study was that reported by Kwallek et al. (1996), but also in this case participants remained in the experimental room only the time to complete some tests for a total duration of about 45 min.
This is the first study that examined color preferences and the effects of environmental color on psychological functioning in a population that lived in mean more than 1 year in an architectural setting characterized by a strong monochromatic color interior design. The innovative aspect of our study was the possibility to examine color preferences and psychological effects of long-term color exposure in a real residential context. The university residence hall provided a setting with a high ecological validity for the study of color influence on residential satisfaction, lightness self-evaluation, study facilitation, and mood.
Participants were 443 university students living in a university residence hall. The sample included 230 males (Mage = 23.91, SD = 2.73) and 213 females (Mage = 23.68, SD = 2.60). The distribution of participants between the six buildings, differing for the specific interior color, was: orange N = 74 (16.7%), blue N = 75 (16.9%), yellow N = 74 (16.7%), red N = 87 (19.6%), green N = 85 (19.2%), and violet N = 48 (10.9%). Student assignment to the different buildings was performed by the residence hall administration at the time of admission. Mean stay at the university residence hall at the time of the research was 13.33 months (SD = 12.14). Difference between mean stay in the six buildings was not significant.
Participants were accommodated in single (31.8%) and double (68.2%) rooms. The proportion of students in single and double rooms was homogeneous for the six buildings. Eight participants were excluded because they declared a deficiency in color vision. Participants declared to spend an average of 6.78 h (SD = 3.28) per day in their room (excluding sleeping time).
This study was carried out in accordance with the recommendations of the Ethics Committee of the University of Bologna that approved the study protocol. All participants gave written informed consent in accordance with the Declaration of Helsinki. The data were collected in an anonymous form.
The study was conducted at the university residence hall “I Praticelli,” located in Pisa (Italy). This setting was chosen for its architectural properties, since the university residence hall is divided into six identical buildings differing only for the interior color (walls, floor, and ceiling) (Figure Figure1 1 ). Each building has common areas (corridors, kitchen, and living roy coordinates are reported in Table Table1 1 . Artificial lighting within the six buildings was uniform with the use of linear fluorescent bulbs (photometrical data: color rendering index Ra ≥ 80, light color 830, rated color temperature 3000K).