Comparative analysis of color emotional perception in art and non-art university students: hue, saturation, and brightness effects in the Munsell color system

Abstract

Background

Hue, saturation, and brightness are critical dimensions of color perception, and art students may experience color-related emotions differently from non-art students due to either their specialized training or innate artistic sensitivity. Despite this intriguing possibility, limited research has systematically examined these differences. This study aims to investigate how art and non-art students perceive and emotionally evaluate different colors.

Methods

Using a questionnaire-based approach, participants rated the emotional valence of colors on a 7-point semantic differential scale, ranging from very negative to very positive. The colors were carefully selected from Munsell's color system, which enabled precise manipulation of hue, brightness, and saturation levels. The study included 36 art students and 36 non-art students, who evaluated colors under controlled experimental conditions.

Results

The findings indicate that brightness and saturation significantly influence emotional responses to colors. Warmer colors, such as orange, elicited more positive emotional evaluations, whereas cooler colors, such as blue, received less positive ratings. Importantly, while differences in emotional responses between art and non-art students were observed, these differences were relatively modest. Art students showed a slight but noticeable sensitivity to variations in brightness and hue, whereas non-art students exhibited stronger responses to color saturation.

Conclusion

The findings reveal measurable differences in how art and non-art students emotionally evaluate color, particularly in response to variations in hue, brightness, and saturation. These results provide new insights into the nuanced emotional interpretation of color across populations, enriching our understanding of affective visual perception.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40359-025-03034-y.

Introduction

The common view is that color plays a pivotal role in visual communication and emotional expression, especially in fields such as art and design [42]. Understanding how individuals perceive and respond to color is essential for enhancing aesthetic experiences and improving communication across various contexts. Three primary dimensions influence color perception: hue (the dominant wavelength), saturation (intensity or purity), and brightness (perceived lightness).

Artists and designers, trained to use color for emotional and communicative purposes, typically exhibit a more nuanced understanding of color perception than the general population [6, 25]. This expertise allows them to anticipate audience reactions and select colors that align with intended emotional effects. Neuroscientific evidence supports these behavioral differences: for example, painting majors show increased activation in the color-selective left V4 area and altered connectivity patterns during color naming tasks, compared to non-artists [25].

Importantly, artistic expertise also appears to shape emotional responses to visual stimuli. Individuals with visual arts training tend to show stronger affective engagement with abstract or semantically ambiguous artworks—stimuli that often elicit weaker reactions in non-artists [13]. These findings suggest that aesthetic sensitivity is influenced by learned top-down mechanisms that modulate not only perceptual strategies but also emotional interpretations.

Despite growing evidence for perceptual and emotional differences between artists and non-artists [13, 25], little is known about how these differences extend to the emotional interpretation of specific color dimensions—such as hue, saturation, and brightness. Given that emotional response is central to color experience in aesthetic contexts, further research is needed to understand how artistic expertise influences the cognitive-emotional integration of color perception.

The emotional impact of color has been extensively studied across psychology and design disciplines [17]. For example, warm colors, such as red and yellow, often evoke feelings of energy and positivity, while cool colors, such as blue and green, are linked to calmness and tranquility [1, 24]. Adams and Osgood [1] found that light colors were perceived as "good" and dark colors as "bad."

Previous studies have examined how color dimensions influence preference and aesthetic appeal. Brighter and more saturated colors are generally rated as more attractive and preferred [2, 5, 7, 18, 40]). However, it is important to distinguish these judgments from emotional responses to color, which reflect a different psychological process. For example, blue is often the most preferred color in terms of liking and aesthetic appeal [30, 36], yet it is frequently associated with sadness in emotional contexts [10].

This study focuses on the perception of emotional valence elicited by color, rather than on preference or attractiveness. Although these dimensions may interact in real-life experiences, they are underpinned by distinct psychological mechanisms. Differentiating among them allows for a more precise understanding of how hue, saturation, and brightness independently contribute to emotional experience.

While previous research has shown that artists and non-artists differ in color appreciation and aesthetic judgment [10, 13], systematic investigations into their perceptual differences across specific color dimensions remain limited. This gap underscores the theoretical and practical significance of comparing emotional responses to color dimensions between groups with and without artistic training.

An expanding body of research supports the notion that artistic training fosters top-down visual perception and spatial reasoning. Visual arts education has been associated with improvements in spatial visualization and analytical observation [19], as well as to more efficient neural processing of visual information in professionally trained artists [6]. Collectively, these findings suggest that sustained artistic training is positively linked to refined perceptual and cognitive visual abilities.

A previous study examined color associations among designers and non-designers, finding significant differences between the two groups in their stereotypical color-concept pairings [28]. For instance, while both groups associated red with concepts such as fire and danger, designers and non-designers differed in their associations with concepts such as potential danger and escape. This underscores the importance of considering background and expertise in color perception, particularly in user-centered design.

The current study aimed to clarify how hue, saturation, and brightness influence emotional interpretations of color by comparing the evaluative and cognitive responses of art and non-art students. We hypothesized that:

• H1: Hue, saturation, and brightness each independently influence emotional valence ratings.

• H2: The effect of hue, saturation, and brightness on emotional perception is moderated by group type, such that the impact of each color dimension differs between art and non-art groups.

Our analyses used a 7-point semantic differential scale to assess momentary emotional reactions to color stimuli [38]. To accurately control color variables, we used the Munsell Color System [8, 21, 22], a widely recognized framework for color classification. This system classifies colors according to three main dimensions: hue, saturation, and brightness. Since failure to control these dimensions may lead to inconsistent or less reliable results, we ensured strict parameterization across all stimuli. Furthermore, participants were divided into a control group with no professional visual arts training and an experimental group with at least three years of art-related experience.

Methods

Participants

A total of 72 participants, including undergraduate and graduate students from Dalian University of Technology, were recruited for this study. The sample comprised 36 participants in the art training group (20 females, 16 males; mean age = 20.81 years, SD = 2.39, range = 18–28 years) and 36 participants in the non-art control group (19 females, 17 males; mean age = 21.91 years, SD = 2.74, range = 20–33 years). All participants completed an online Ishihara color blindness screening test [26] to confirm normal color vision and reported no history of color deficiency.

Prior to data collection, an a priori power analysis was conducted using G*Power 3.1.9.7 [14, 32] to determine the minimum required sample size. The analysis was based on a repeated measures ANOVA with a within–between interaction, given the study's mixed design (within-subjects factors: brightness, saturation, hue,between-subjects factor: art training). Assuming a medium effect size (Cohen’s f = 0.25), an alpha level of 0.05, and a desired power of 0.95, the analysis indicated that 27 participants per group (total N = 54) would be sufficient. These parameters were based from prior studies on the color–emotion associations [37]. No pilot data were used in the estimation. The final sample size of 72 participants exceeded this threshold, suggesting the study was adequately powered to detect medium-sized effects.

Participants in the art training group were students majoring in art or design, with an average of 6 years of specialized art education (M = 6.00 years, SD = 3.28, range = 2–13 years). Participants in the non-art control group had no formal or amateur art training experience. This group distinction was critical for examining differences in color perception between individuals with and without artistic training.

Online experiment

The online experiment was organized into three sections, each dedicated to one of the fundamental dimensions of color: brightness, saturation, and hue. The experiment was administered through the Microsoft Forms platform, a widely used online survey tool. Participants first completed a brief pre-test to verify monitor contrast and display settings. After the pre-test, they proceeded to the main experiment, which was divided into three sequential blocks (brightness, saturation, hue). In each block, participants rated the emotional valence of a series of color stimuli on a 7-point semantic differential scale, where 1 = very negative, 2 = moderately negative, 3 = slightly negative, 4 = neutral, 5 = slightly positive, 6 = moderately positive, and 7 = very positive. No time limits were imposed, but participants were required to provide a response before moving on to the next item.

To ensure chromatic consistency, the color stimuli were carefully selected from the 10 primary hues in the Munsell color system, with a fixed chromaticity level of 5 for each hue. This specific chromaticity was chosen because it represents the most prototypical or characteristic color within each hue category. The hues included purple-blue (PB), blue (B), blue-green (BG), green (G), green-yellow (GY), yellow (Y), yellow–red (YR), red (R), red–purple (RP), and purple (P).

The Munsell color system, a framework rooted in psychological principles of visual perception, provided the basis for the selection of color stimuli. An applet was employed to randomly select 10 colors for each of the 10 primary hues. Randomization played a crucial role in the experimental design, helping to control variables and reduce potential confounding factors [4].

Each color stimulus was presented as a square patch occupying approximately 1024 × 798 pixels on participants'screens. To minimize display variability, participants were instructed to complete the experiment using a laptop or desktop computer (rather than a mobile phone or tablet), ensure that any screen filters such as night mode or blue light reduction were turned off, and work in a well-lit environment free from visual distractions. Additionally, a brief pre-test was included to ensure screen adequacy, where participants identified grayscale contrast differences before proceeding.

Brightness stimuli selection

To investigate the effect of brightness on color perception, two sets of color stimuli were selected for each of the 10 primary hues from the Munsell Color System, with brightness being the only manipulated variable. Within each pair, the stimuli differed by six levels of brightness while keeping hue and saturation constant. This resulted in a total of 20 brightness-manipulated stimuli (10 pairs). Figure 1b provides an example from the 5YR hue, showing the high- and low-brightness stimuli.

Fig. 1.

a The Munsell color system represented in cylindrical coordinates. b Illustration of the method for selecting two colors of different brightness levels using the orange hue as an example. c Illustration of the method for selecting two colors of different saturation levels, also using the orange hue as an example. d Randomly selected color samples from the 10 main Munsell hue categories. Note: Subfigure (a) is adapted from Munsell 1943 color solid cylindrical coordinates, Wikimedia Commons (https://commons.wikimedia.org/wiki/File:Munsell_1943_color_solid_cylindrical_coordinates.png), licensed under CC BY-SA 4.0. Subfigures (b–d) were created by the authors to illustrate the experimental procedure

Saturation stimuli selection

For the saturation study, two levels of saturation (high and low) were chosen for each of the 10 primary hues, while brightness and hue were held constant. This design allowed for direct comparison between saturation levels without confounding influences from other dimensions. A total of 20 saturation-manipulated stimuli (10 pairs) were used. Figure 1c illustrates an example from the 5YR hue.

Hue stimuli selection

In the hue study, a total of 100 color stimuli were selected—10 colors from each of the 10 primary hues in the Munsell Color System. These colors were randomly sampled using a custom-coded Matlab applet. Randomization was a key component of the experimental design, aiming to minimize potential confounding factors such as variations in lightness and saturation across hues [4]. Each color sample within a hue was assigned a numeric code, and the applet randomly selected 10 colors per hue while controlling for brightness and saturation. This approach ensured that the final color set was both representative and comparable across hue categories. Figure 1d displays the selected color stimuli from the 10 primary hues.

Based on established principles in color perception and wavelength theory, the 10 hues in this study were categorized into warm and cool colors for further analysis. Warm colors (Y, YR, R, RP, P) are generally associated with longer wavelengths (~ 570–750 nm), while cool colors (PB, B, BG, G, GY) correspond to shorter wavelengths (~ 450–570 nm).

This categorization is not arbitrary but rooted in both physiological mechanisms and perceptual-emotional associations. Human vision relies on three types of cone photoreceptors, each maximally sensitive to different ranges of wavelengths—long (L), medium (M), and short (S). Long-wavelength light (e.g., red) tends to be perceived as closer and warmer, while short-wavelength light (e.g., blue) is perceived as distant and cooler due to the spatial and depth-related cues processed by the visual system [40].

Moreover, the structure of the Munsell system reflects these perceptual differences. Designed for perceptual uniformity, it shows that warm hues occupy a broader chroma range than cool hues [44].This reflects the human eye’s greater sensitivity to longer wavelengths, enabling finer chromatic discrimination in warm color regions [8]. As shown in Fig. 1d, warm hues (Y, YR, R, RP, P) include more chromatic steps in the Munsell solid, whereas cool hues (PB, B, BG, G, GY) are represented with fewer chromatic variations.

These distinctions are also reflected in affective responses: warm hues are often linked with higher arousal and emotional intensity, while cool hues tend to evoke calmness and relaxation [23, 29, 37]. The warm–cool dichotomy is widely applied in visual arts, interior design, and color psychology to model emotional and spatial reactions to color [10, 17].

Statistical analysis

The study design incorporated art training as a between-subjects factor (art vs. non-art), while brightness (high vs. low), saturation (high vs. low), and hue (10 levels) were treated as within-subjects factors across the three studies. For the hue analysis, hues were further grouped into warm (Y, YR, R, RP, P) and cool (PB, B, BG, G, GY) categories based on the physical properties of light, with warm colors corresponding to longer wavelengths and cool colors to shorter wavelengths.

A repeated-measures ANOVA (rm-ANOVA) was conducted to examine the main effects and interaction effects of group and each color attribute. When Mauchly’s test indicated a violation of sphericity, degrees of freedom were corrected using the Greenhouse–Geisser adjustment to maintain the validity of the F-test. For each ANOVA, the F-value, associated p-value, and partial eta squared (${\eta }_p^2$) are reported as indices of statistical significance and effect size.

To further explore significant interaction effects, simple effects analyses were conducted using paired-samples t-tests. For these comparisons, Cohen’s d was calculated to quantify the magnitude of differences, and 95% confidence intervals (CIs) for the difference estimates were reported to indicate the precision of the effect sizes, following best practices by Gignac and Szodorai [15]. All statistical tests were two-tailed, with the significance threshold set at α = 0.05. Statistical analyses were performed using IBM SPSS Statistics 26.

Results

Effect of brightness on emotional scores

The statistical analysis revealed a significant effect of brightness level, F(1, 70) = 431.295, p < 0.001, ${\eta }_p^2$ = 0.860, as well as a significant interaction between group and brightness, F(1, 70) = 5.426, p = 0.023, ${\eta }_p^2$ = 0.072.

Simple effects analyses indicated that both the art group (t(35) = 18.948, p < 0.001, d = 3.158, 95% CI [2.330, 3.986]) and the non-art group (t(35) = 11.223, p < 0.001, d = 1.871, 95% CI [1.309, 2.432]) rated high- and low-brightness colors significantly differently (see Fig. 2b).

Fig. 2.

a Between-group comparison of emotional valence ratings for different brightness levels. b Within-group comparison of emotional valence ratings for different brightness levels. c Between-group comparison of emotional valence ratings for different saturation levels. d Within-group comparison of emotional valence ratings for different saturation levels. e Emotional valence ratings across 10 hue categories for art and non-art groups. f Between-group comparison of emotional valence ratings across hues. g Within-group comparison of emotional valence ratings across hues. Note: All data points and connecting lines represent group means; error bars indicate standard deviations (SD)

In addition, a marginally significant difference was found between the two groups for color mood scores under high brightness (t(35) = 1.948, p = 0.059, d = 0.435, 95% CI [− 0.016, 0.760]). While this result did not reach the conventional threshold for statistical significance (p < 0.05), it suggests a trend toward higher ratings by art students.

Effect of saturation on emotional scores

The analysis also revealed a significant main effect of saturation level, F(1, 70) = 199.485, p < 0.001, ${\eta }_p^2$ = 0.740, and a significant group × saturation interaction, F(1, 70) = 7.655, p = 0.007, ${\eta }_p^2$ = 0.099, see Fig. 2c.

Simple effects analyses further showed a significant group difference under the high-saturation condition (t(35) = 2.307, p = 0.027, d = 0.385, 95% CI [0.034, 0.735]).

Additionally, both the art group (t(35) = 8.005, p < 0.001, d = 1.334, 95% CI [0.869, 1.799]) and the non-art group (t(35) = 11.295, p < 0.001, d = 1.883, 95% CI [1.319, 2.446]) perceived significant emotional differences between high and low saturation colors (see Fig. 2d).

Effect of hue on emotional scores

The statistical analysis (Fig. 2e) revealed a significant main effect of hue, F(1, 70) = 43.774, p < 0.001, ${\eta }_p^2$= 0.385, as well as a significant interaction between group and hue, F(1, 70) = 2.532, p = 0.032, ${\eta }_p^2$ = 0.035.

Among the ten tested hues, yellow and orange received the highest emotional ratings, whereas blue-violet and blue were rated the lowest (see Fig. 2e and Table 1). Based on wavelength characteristics, the ten hues were categorized into warm and cool colors for further analysis (cool colors: PB, B, BG, G, GY; warm colors: Y, YR, R, RP, P).

Table 1.

Results of emotional valence ratings for the 10 hue colors by participants in both groups

	Art Group		Non-art Group
Hue	Mean ± SD	95% CI	Mean ± SD	95% CI
P	3.917 ± 0.102	3.711–4.123	4.041 ± 0.121	3.795–4.288
PB	3.521 ± 0.095	3.329–3.713	3.778 ± 0.131	3.513–4.043
B	3.688 ± 0.102	3.455–3.920	3.882 ± 0.122	3.634–4.130
BG	4.097 ± 0.116	3.851–4.343	4.285 ± 0.141	3.999–4.571
G	4.063 ± 0.095	3.869–4.256	4.188 ± 0.122	3.940–4.436
GY	4.444 ± 0.112	4.217–4.672	4.472 ± 0.125	4.218–4.726
Y	4.924 ± 0.093	4.735–5.112	4.778 ± 0.121	4.533–5.023
YR	4.889 ± 0.107	4.673–5.105	4.611 ± 0.106	4.397–4.826
R	4.604 ± 0.107	4.386–4.822	4.410 ± 0.119	4.168–4.652
RP	4.361 ± 0.116	4.126–4.596	4.229 ± 0.122	3.982–4.476

PB Purple-blue, B Blue, BG Blue-green, P Purple, G Green, GY Green-yellow, Y Yellow, YR Yellow–red, RP Red–purple, R Red

The follow-up analysis showed a significant main effect of color temperature (warm vs. cool), F(1, 70) = 117.632, p < 0.001, ${\eta }_p^2$ = 0.627, and a significant interaction between group and hue category, F(1, 70) = 7.782, p = 0.007, ${\eta }_p^2$= 0.100.

Simple effects analysis further revealed that participants in the art group perceived a significant difference between warm and cool colors (t(35) = 9.667, p < 0.001, d = 1.611, 95% CI [1.098, 2.124]; see Fig. 2g.

Discussion

In this study, we employed the Munsell color system to manipulate the three core color dimensions—brightness, saturation, and hue—and examined their impact on emotional perception. Additionally, we investigated differences in emotional color perception between art and non-art students, having both groups rate the emotional tone of various color stimuli.

The statistical analysis results revealed several important insights. First, colors with higher brightness and saturation were rated more positively than those with lower brightness and saturation. Overall, warm colors were perceived as more positive than cool colors, with orange rated as the most positive and blue as the least positive. These findings lend support to H1, demonstrating that brightness, saturation, and hue each significantly affect emotional responses. Moreover, differences in color perception between the two groups were observed: art students rated colors with high brightness and warm tones more positively, while non-art students responded more favorably to highly saturated colors. This pattern supports H2, suggesting that art and non-art students differ in their emotional responses to color. This may be due to perceptual training or innate differences between the groups, reflecting either the influence of artistic education or pre-existing individual traits.

Both groups in our study rated warm colors (e.g., red, orange, yellow) more positively, highlighting a shared affective association with warmth, energy, and joy. This pattern may reflect broadly shared cultural or psychological associations, where warm colors are often linked to feelings of energy, joy, and excitement. Supporting this, prior studies have consistently found that yellow and orange are frequently associated with happiness and warmth, while red is linked to high arousal emotions such as excitement and passion [17, 23].

However, the two groups still demonstrated different sensitivities to color dimensions: art students responded more positively to high-brightness and warm colors, whereas non-art students showed greater positivity toward high-saturation colors. Understanding these perceptual and emotional differences is critical for user-centered design, where color choice can significantly impact emotional responses.

The differences observed between art and non-art students corroborate findings from Long et al. [25] and Damiano et al. [10], although the small effect sizes indicate modest effects that should be interpreted with caution. These group differences were subtle and should not be regarded as strong evidence of a causal influence of artistic training; rather, they may reflect pre-existing perceptual tendencies or self-selection bias. Future work should directly measure training-specific variables (e.g., years and intensity of instruction) to disentangle learning effects from innate abilities.

Moreover, these results align with Mariko Shirai's [38] conceptual framework for emotional measurement, validating the use of the 7-point semantic differential scale to evaluate emotional associations with color. However, it is important to note that this scale captures momentary emotional responses rather than more sustained mood states. Future research could benefit from employing additional measures to directly assess mood changes resulting from exposure to different colors.

In line with previous research, our findings confirm that brightness significantly influences emotional perception. For example, Chen et al. [7] found that brighter colors tend to be perceived as more attractive and emotionally appealing, which aligns with our results. Similarly, Park et al. [31] reported that participants experienced more positive emotions in brighter visual conditions. In our study, art students exhibited slightly greater sensitivity to brightness levels compared to non-art students, possibly reflecting perceptual enhancements cultivated through sustained artistic training. While the difference was modest, a marginal group effect for high-brightness stimuli (p = 0.059, d = 0.435) suggests a potential trend toward greater positivity among art students, potentially indicating subtle training-related modulation in affective color processing.

Moreover, the observed tendency for art students to rate warm colors (e.g., red, orange, yellow) more positively than non-art students may reflect mechanisms shaped by artistic training. Warm hues are commonly used in visual art to express emotional intensity, vitality, and human-centered themes such as warmth and passion [12].Repeated exposure to these symbolic functions may increase art students’ sensitivity to the emotional qualities of such colors, leading to more positive valence evaluations.

From a cognitive-perceptual perspective, artistic training may enhance top-down modulation in visual processing [41], promoting deeper semantic associations between color and emotion. This supports expertise theories, which suggest that experts develop more nuanced and affect-rich interpretations of domain-relevant stimuli [16]. In this context, warm hues—frequently used in expressive compositions—may hold greater emotional salience for art students than for their non-art peers.

Such perceptual differences may also manifest at the neural level, as recent findings suggest. For instance,Song et al., [39]reported significant group differences (art vs. non-art students) in neural responses—specifically in the N2 and P3 components of event-related potentials (ERPs)—when participants viewed colors with varying brightness levels. The N2 component is commonly associated with attentional control and conflict monitoring, whereas the P3 component reflects stimulus evaluation and categorization. Art students exhibited greater N2 and P3 amplitudes, suggesting enhanced top-down modulation and more refined visual processing, even in the absence of observable behavioral differences. In contrast to the 3-point scale used in their study, our adoption of a 7-point semantic differential scale likely enabled more fine-grained and sensitive assessments of emotional valence [11].

Saturation also played an important role in influencing emotional perception, consistent with previous studies [5, 37, 43]. Highly saturated colors were generally associated with more positive emotions. However, in our study, non-art students tended to rate highly saturated colors more positively than art students did, suggesting differing aesthetic preferences for colors between the two groups. Reymondet al., [33]proposed that color saturation affects laypeople and experts differently: while increased saturation may enhance positive evaluations for laypeople, it can induce confusion or overstimulation for experts. This aligns with our findings, where art students exhibited more restrained emotional responses to saturation.

In contrast, art students were more sensitive to brightness differences, rating high-brightness stimuli more positively than their non-art counterparts. This divergent pattern raises an important question: why does artistic training appear to enhance sensitivity to brightness but attenuate emotional responses to saturation? One possibility is that brightness is a more fundamental perceptual cue in the organization of form, contrast, and spatial hierarchy in visual art, and as such, it is extensively attended to during formal training. Brightness is also more closely linked to luminance perception, which artists frequently manipulate in shading, tonal balance, and atmospheric effects. In contrast, saturation may be treated with more caution in art education, where overly saturated colors are often viewed as garish or unsophisticated unless handled deliberately. Thus, artistic training may promote selective refinement of perceptual-emotional associations based on the functional and aesthetic priorities of the visual arts discipline. These findings suggest that art training may not uniformly modulate emotional color perception but may instead produce domain-specific shifts in perceptual sensitivity, depending on how particular color dimensions are used and valued within artistic practice.

Different hues evoke distinct emotional responses. According to Russell’s circumplex model of affect [34, 35], emotions can be described along two primary dimensions: valence (pleasant–unpleasant) and arousal (high–low activation). Warm colors, such as red and orange, have often been linked to feelings of comfort, enthusiasm, and passion, whereas cool colors, such as blue, evoke calmness and relaxation [20].

Our findings suggest that warm hues tend to receive more positive emotional ratings in both groups, whereas cool hues are associated with emotionally neutral or mildly negative valence. This pattern aligns with previous research [3, 9], which indicates that warm colors often evoke high-valence, high-arousal emotions such as excitement or enthusiasm. However, they may also elicit negatively valenced high-arousal states such as anger [35].In contrast, cool hues like blue are typically linked to lower arousal levels and are more frequently associated with negative low-arousal emotions such as sadness, as well as calm and peaceful emotional states. These differences underscore the importance of considering both valence and arousal dimensions in understanding emotional responses to color. Notably, art students rated warm hues more positively than non-art students, potentially reflecting enhanced color sensitivity developed through artistic training—though the magnitude of group differences was smaller than initially anticipated.

While this study provides valuable insights, several limitations warrant consideration. First, it remains unclear whether the observed differences arise from artistic training or inherent talent—individuals with superior perceptual or cognitive abilities may self-select into art disciplines, potentially confounding the results [6]. Second, the experiment was conducted online to facilitate broader and more efficient recruitment, especially for art students who are typically difficult to reach through traditional, in-person methods. the questionnaire-based approach precluded real-time tracking of participants’ responses, which may affect data reliability and temporal resolution While this approach enhanced participant access, it also introduced variability in color presentation across different display devices. This limitation reflects the inherent trade-off in web-based color research between ecological validity and perceptual control, rather than any shortcomings of the Munsell system itself.

Finally, although we focused solely on valence ratings to assess emotional responses to color, emotional experience is inherently multidimensional. Established frameworks such as the PAD (Pleasure–Arousal–Dominance) model emphasize arousal and dominance as key dimensions alongside valence [27]. Limiting our assessment to valence may have constrained the depth and nuance of emotional interpretation. Future research should incorporate multidimensional emotion models, integrating arousal and dominance ratings as well as physiological measures (e.g., skin conductance, heart rate variability), and examine training-specific variables to provide a more comprehensive understanding of affective responses to color and the potential role of artistic experience.

Conclusion

This study found that brightness and saturation significantly influenced emotional ratings, with higher levels generally linked to more positive responses. Warm hues were rated more positively than cool hues, especially by art students, who also showed greater sensitivity to brightness changes, while non-art students responded more to saturation. These findings deepen our understanding of how color dimensions—brightness, saturation, and hue—shape emotional perception. While art students exhibited some distinct response patterns, it remains unclear whether these reflect the effects of artistic training or pre-existing individual differences. Thus, the potential influence of art education should be interpreted with caution. Overall, our results highlight the complex interplay between perceptual features and individual background in shaping affective responses to color.

Authors’ contributions

All authors were involved in the study design, discussion of the results, and writing of the final manuscript. L.S. and L.M. designed the questionnaire, L.S. collected the questionnaires and conducted the study with the support of F.C., L.S., and G.Z. analyzed the data, L.S. wrote the manuscript, F.C., G.Z. and J.S. made several revisions.

Funding

Scholarships from the China Scholarship Council (No. 201906060241).

Data availability

The datasets generated and/or analyzed during this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Medical Engineering Ethics Review Group of Dalian University of Technology, China, and conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants prior to their involvement in the study, ensuring their understanding and willingness to participate.

Consent for publication

The findings presented in this manuscript are original and have neither been published elsewhere nor submitted for consideration to any other publisher by any of the authors.

Competing interests

The authors declare no competing interests.