Effect of Color–Shape Associations on visual feature discrimination

doi:10.21203/rs.3.rs-2072956/v1

Download PDF

Research Article

Effect of Color–Shape Associations on visual feature discrimination

https://doi.org/10.21203/rs.3.rs-2072956/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Non-synesthetes exhibit a tendency to associate specific shapes with particular colors (i.e., circle–red, triangle–yellow, and square–blue). Here, we used two Go/No-go tasks to examine the congruency priming effect of color-shape associations on recognition efficiency of color and shape features. At the beginning of each trial, a target color or shape word was introduced, followed by a colored-shape visual stimulus. Participants were required to press a key to a target stimulus (“go” cues), while withholding their responses to a non-target stimulus (“no-go” cues). The targets were presented either visually (visual word, Experiment 1) or auditorily (spoken word, Experiment 2). Results showed a congruency effect of color-shape associations on recognition efficiency for color and shape features in both experiments. Response times were shorter in congruent than incongruent conditions, that a target could be recognized faster when it was presented with the congruent visual features than with incongruent ones, irrespective of the presentation form (visual or auditory). These results suggest that color-shape associations can be strong to influence visual recognition of color and shape features.

Psychology

What color best matches a circle? When asked to choose a color that best matches a shape, people typically show a systematic pattern of color-shape associations (CSAs); that is, red for a circle, yellow for a triangle, and blue for a square (Chen et al., 2015b). The study of color-shape associations can be traced back to the famous abstract painter Wassily Kandinsky (Kandinsky, 1912, 1947). He firstly claimed that there was an inherent association between colors and shapes, such that a circle is best matched with blue, a square is red, and a triangle is yellow. However, it evoked strong opposition and debate from fellow artists and researchers, and modern empirical findings suggested different patterns of CSAs (e.g., Albertazzi et al., 2013; Chen et al., 2015a, 2015b; Dumitrescu, 2011; Hanada, 2019; Jacobsen, 2002; Jacobsen & Wolsdorff, 2007; Kharkhurin, 2012; Lupton & Miller, 1993). Chen et al. (2015b) using an explicit color-shape matching task observed that Japanese people systematically associated specific shapes with particular colors (i.e., circle-red, square-blue, and triangle-yellow), which were consistent with Italian participants (Albertazzi et al., 2013). They firstly proved that those CSAs could be strong to test by both direct (Chen et al., 2015b) and indirect experimental methods (e.g., implicit association test [IAT]; Chen et al., 2015a). Chen et al (2015b) further proved that common semantic information between colors and shapes (e.g., warm/cold) could explain these associations. Thus, CSAs would appear to be acquired rather than innate. Moreover, statistical (Jacobsen, 2002; Jacobsen & Wolsdorff, 2007; Hanada, 2019) and emotional (Malfatti, 2014; Dreksler & Spence, 2019) correspondence hypotheses were also suggested to account for CSAs. For instance, Dreksler and Spence (2019) suggested that the preferences and arousal appraisals for colors and shapes influence those CSAs. Hanada (2019) suggested that CSAs might be learned from environment (e.g., circle-red from the Japanese flag).

To date, most studies used directive matching tasks to examine CSAs, which allowed participants referring to their stored knowledge and experience with colored-objects. Few studies have examined the congruency effect of CSAs on visual processes by indirect behavioral tasks. Makin and Wuerger (2013), and Chen et al. (2015a) adopted the implicit association tests (IATs) to examine Kandinsky’s CSAs in British and Japanese participants, respectively. Both showed limited supports for the congruency effect of Kandinsky’s CSAs on response times. Chen et al. (2015a) also used IATs to test Japanese CSAs (i.e., circle-red, square-blue, and triangle-yellow). They observed a robust congruency effect of CSAs across three IAT tasks, that response times were significantly faster when congruent color-shape mapping onto the same response fingers than incongruent color-shape mapping onto the same response fingers. Thus, CSAs could be strong enough to influence response times for recognizing color and shape features. It may be noticed that IAT reveals associations through faster response times and more accuracy when associated concepts share the same behavior responses (Greenwald et al., 1998). Thus, IAT measures the relative strength of associations involving two pairs of categories. However, IATs may not distinguish which association of the two categories drives the congruency effect, and little has been known on the priming effect of congruency CSAs on visual feature discrimination. The Go/No-go task has an advantage over the IAT, that it can assess the strength of implicit associations between a single target category and an attribute dimension (Nosek & Banaji, 2001). In a Go/No-go task, participants are required to press a key if the presented stimulus belongs to the target category (go signal), where the association between those categories is being assessed, and to not respond if the stimulus is a non-target (no-go signal). If there is a stronger association between the target and attribute, participants should distinguish the signals more readily and make less errors. Thus, Go/No-go task can provide further insight into the strength of CSAs influencing visual recognition with colors and shapes, such as the magnitude of interference between those two dimensions (Holmes & Zanker, 2013; Kharkhurin, 2012; Makin & Wuerger, 2013).

Based on previous findings, the current study aimed to explore the congruency effect of CSAs (i.e., circle-red, triangle-yellow, and square-blue) on visual feature discrimination for colors and shapes. In a more exploratory manner, we tested the congruency effect of CSAs with modality manipulation using both visual word and auditory word input, to examine whether CSAs could be elicited irrespective of the target input modality. We hypothesized that a congruent CSAs priming will improve participants’ performance in feature discrimination of colors and shapes compared with incongruent priming conditions in both visual and spoken word forms.

2.1 Method

2.1.1 Participants

Thirty-one Japanese college students participated in this study (17 females, mean age = 20.7 years, SD = 1.7). All participants had normal or corrected-to-normal visual acuity and normal color vision, normal motor functions, and they were naïve to the study purpose. Sample size was determined using prior experiments investigating similar effect (Cohen’s d_z = 0.5; Velasco et al., 2015). This experiment was approved by the institutional review board of Waseda University (2015-033) and conducted in accordance with the ethical standards of the 1964 Declaration of Helsinki. Written informed consent was obtained from all participants in advance.

2.1.2 Apparatus and stimuli

E-Prime 2.0 software (Psychology Software Tools, Inc.; http://pstnet.com/products/e-prime/) was used to present the stimuli and collect the data. Stimuli were presented on a 24-inch LCD monitor (EIZO FG2421, EIZO Corp., Hakusan, Japan), with a 1920 × 1080-pixel resolution and a refresh rate of 100 Hz.

Six Japanese kanji characters and nine colored shapes (three colors—red, yellow, and blue—combined with three shapes—circle, triangle, and square) were used as visual stimuli. The target words: “円 (Circle),” “三角 (Triangle),” “四角 (Square),” “赤 (Red),” “黄 (Yellow),” and “青 (Blue)” were presented in a 40-point MS Mincho font. The colored-shape stimuli were all white edge line drawings on a black background (8 cd/m²). The circle was 2.8° in diameter, the square was 2.4° (in height) × 2.4° (in width), and the triangle was 2.8° (in height) × 3.2° (in width). The stimuli were always presented in an upward orientation and centered on the screen (see Fig. 1). The colored-shape stimuli were measured using PR-655 (Photo Research, Chatsworth, CA, USA) and each color was consecutively measured 10 times and averaged. The color information was as follows: yellow: L* = 124.1, a* = -27.3, b* = 131.5; red: L* = 73.6, a* = 83.1, b* = 101.5; and blue: L* = 76.1, a* = 11.9, b* = -103.0.

2.1.3 Procedure

The experiment was conducted in a dimly lit room. Each participant was seated about 60 cm in front of the monitor. At the beginning of the task, written instructions were provided on the computer screen, followed by the practice trials. In each trial, participants were shown a word target in Japanese kanji character for 1000 ms, followed by a central fixation cross (for 300 ms), and then a colored-shape stimulus (for a maximum of 1500 ms). Trials were separated by an inter-stimulus interval (ISI) of 300 ms. Participants were instructed to press the space key with their index finger as rapidly as possible if the colored-shape stimulus belonged to the target word category (“go” trial). For a congruent “go” trial, the colored-shape stimulus was consistent with CSAs (i.e., circle-red, triangle-yellow, and square-blue); however, for an incongruent “go” trial, the colored-shape was aligned with the target category, but inconsistent with CSAs. For example, for the “circle (円)” word target, a red-circle stimulus made a congruent “go” signal, and a blue-circle or yellow-circle made an incongruent “go” signal. Participants were instructed to withhold responding if the colored-shape stimulus did not belong to the target category (“no-go” trial). For the “circle” target, a colored-square or colored-triangle made a “no-go” signal (see Table 1 for the complete list; Fig. 1). We aimed to compare the response times (RTs) in congruent-go and incongruent-go trials, therefore, we used the same number of trials in congruent-go, incongruent-go, and no-go conditions (Schulz et al., 2007).

In order to equal the trial numbers of the three conditions, we adopted six congruent-go trials (e.g., for circle target, red-circle appeared for six times), six incongruent-go trials (for circle target, yellow-circle and blue circle appeared three times each), and six no-go trials for each target (for circle target, each colored-shape stimulus appeared once; see Table 1). All unique trials were presented twice, giving rise to 36 trials per target and 216 trials for the main experiment, (6 + 6 + 6) × 2 (repeat times) × 6 (targets) = 216 trials. The experiment started with 20 practice trials. Feedback was given during practice, making the participants fully aware of the expected classification. The main 216 trials were broken into six blocks of 36 trials. The trials were randomly presented. At the end of each block, participants had a self-timed break. The whole experiment lasted about 15 mins.

Table 1

Visual stimuli in the three conditions
	Circle	Triangle	Square	Red	Yellow	Blue
Congruent “Go” signals	Red-circle	Yellow-triangle	Blue-square	Red-circle	Yellow-triangle	Blue-square
Incongruent “Go” signals	Yellow-circle/ blue-circle	Red-triangle/ blue-triangle	Red-square/yellow-square	Red-triangle/ red-square	Yellow-circle/ yellow-square	Blue-circle/ blue-triangle
Target absent “No-go” signals	Colored-triangle/ colored-square	Colored-circle/ colored-square	Colored-circle/ colored-triangle	Yellow-shape/ blue-shape	Red-shape/ blue-shape	Red-shape/ yellow-shape

2.1.4 Data analysis

To examine the congruency effect of CSAs on discrimination efficiency for color and shape features, mixed-effect modeling was conducted with repones time (RT) as dependent variable. Data analyses were performed with R 4.0.0 software (R Core Team, 2020). The lmer() function of the “lme4” package (Bates et al., 2015) was used for mixed-effect regression models. We adopted this approach because a mixed-effect model depicts the response as a combination of fixed and random effects, allowing for variation in effects based on random factors such as participants and targets. The main interest is the effect of congruency condition (congruent vs incongruent) on RTs, thus, the fixed factor was congruency condition, whereas random factors were intercepts of subjects and targets or target categories. F and p values were estimated using Satterthwaite approximations using anova() function in the “lmerTest” package (Kuznetsova et al., 2017). The predictors’ coefficients (β-value), the SE-value, t-value (Satterthwaite's method) and the associated p-values generated from each model were reported. In addition, to provide further evidence, RTs in congruent and incongruent conditions were compared using Paired-sample t-tests with Bonferroni correction (α = 0.05/2 = 0.025) for color and shape target categorizes. To further quality the results, the Bayes Factors (BF10) were referred to, in order to determine whether or not there was support in favor of the alternative (H1) or null (H0) hypotheses (Morey & Rouder, 2018). A value of 1 means that null and alternative are equally likely, larger values suggest that the data are in favor of the alternative hypothesis, and smaller values indicate that the data are in favor of the null hypothesis.

2.2 Results

One participant was excluded from data analysis for having exceeded 2SD in either RT or accuracy. Error trials (1.2%) were excluded from data analysis. RTs that fell 2 SDs above or below the mean of each participant were excluded from analysis (4.8% of the remaining trials).

A linear mixed-effect model was fitted to the data with participants and target words as random variables and the congruency condition as the fixed factor (RT ~ Congruency condition + (1|Target word) + (1|Subject)) to reveal the main effect of congruency conditions. Result showed a significant effect of congruency condition on RTs, β = 7.77, SE = 2.09, t = 3.73, p = .0002. Participants responded faster in congruent (mean RT = 385.80 ms, SD = 78.78) than incongruent conditions (mean RT = 393.48 ms, SD = 85.45).

In order to reveal the interaction between congruency condition and target category (color and shape), mixed-effect model using the interaction between congruency condition and target category as fixed factors was fitted to the data (RT ~ Congruency condition + Target category + Congruency condition*Target category + (1|Subject)). Results showed no significant interaction effect between congruency condition and target category, β = 1.08, SE = 4.18, t = 0.26, p = .80. The interaction effect was removed from the model analysis. A mixed-effect model using congruency condition and target category as fixed factors (RT ~ Congruency condition + Target category + (1|Subject)) showed a significant main effect of target category, β = 9.58, SE = 2.96, t = 3.24, p = .001. Participants responded faster for recognizing color targets (mean RT = 384.92 ms, SD = 81.91) than shape targets (mean RT = 394.44 ms, SD = 82.38). Further paired-sample t-tests showed significant effect of congruency condition on RTs for both color and shape targets. For color targets, participants responded faster in congruent (mean RT = 381.38 ms, SD = 77.11) than incongruent conditions (mean RT = 388.46 ms, SD = 86.33; t(29) = 2.96, p = .006, Cohen’s d = 0.14, BF₁₀ = 6.93; Fig. 2). For shape targets, participants responded faster in congruent (mean RT = 390.31 ms, SD = 80.23) than incongruent conditions (mean RT = 399.54 ms, SD = 84.29; t(29) = 3.35, p = .002, Cohen’s d = 0.16, BF₁₀ = 16.15; Fig. 2).

Experiment 1 demonstrated that the congruency priming effect of CSAs influenced recognition efficiency for visual color and shape features. In Experiment 2, we tested whether we could replicate and extend these congruency effect of CSAs using an auditory priming effect of spoken word input. The visual words of color and shape targets in Experiment 1 was changed into spoken words input in Experiment 2, and the other experimental settings were identical.

3.1 Method

3.1.1 Participants

Thirty-four newly recruited individuals participated (16 females, mean age = 21.3 years, SD = 2.5). All participants had normal or corrected-to-normal visual acuity and normal color vision, normal motor functions, and they were naïve to the study purpose. This experiment was approved by the institutional review board of Waseda University (2015-033) and conducted in accordance with the ethical standards of the 1964 Declaration of Helsinki. Written informed consent was obtained from all participants in advance.

3.1.2 Apparatus and Procedure

The apparatus and experimental procedure were identical to those in Experiment 1, except for the visual word stimuli target used in Experiment 1, which was replaced with the spoken sound of each word target. The auditory words were spoken by a female voice:まる (maru; circle), さんかく (sankaku; triangle), しかく (shikaku; square), あか (aka; red), きいろ (kiiro; yellow), and あお (aoi; blue). In order to balance the length of spoken sound for all the color and shape features, we used “maru (circle)” to represent the circle shape instead of “en (circle)”. The six sound files were download from the website: https://soundoftext.com. The sounds were presented over closed-ear headphones (Loudness range = 64–74 dB SPL). At the beginning of the task, written instructions were provided, followed by the practice trials. In each trial, a central fixation cross was presented for 300 ms; then, a word target was presented through the headphones for 1000 ms, following by a visually displayed colored-shape stimulus (for a maximum of 1500 ms). Trials were separated by an ISI of 300 ms. Participants were instructed to press the space key with their index finger as rapidly and accurately as possible if the later presented visual colored-shape stimulus belonged to the spoken target category (either a congruent or incongruent “go” trial) and withhold from responding if the colored-shape stimulus did not belong to the sound target category (“no-go” trial).

3.2 Results

We excluded error trials (0.93%) from data analyses. RTs that fell 2 SDs above or below the mean were excluded from analysis (4.83% of the remaining trials). A linear mixed-effect model was fitted with participants and target words as random variables and the congruency condition as the fixed factor (RT ~ Congruency condition + (1|Target word) + (1|Subject)). Result showed a significant effect of congruency condition on RTs, β = 7.13, SE = 1.92, t = 3.71, p = .0002. Participants responded faster in congruent (mean RT = 397.99 ms, SD = 82.60) than incongruent condition (mean RT = 405.05 ms, SD = 85.14).

Mixed-effect model using the congruency condition (congruent, incongruent), target category (color, shape), and the interaction between congruency condition and target category as fixed factors (RT ~ Congruency condition + Target category + Congruency condition*Target category + (1|Subject)) showed a significant interaction effect between congruency condition and target category, β = 8.68, SE = 3.87, t = 2.25, p = .025. Further analysis showed significant main effect of congruency condition on RTs for spoken words of shape targets, that participants responded faster in congruent (mean RT = 403.57 ms, SD = 86.31) than incongruent conditions (mean RT = 415.37 ms, SD = 92.27; t(33) = 4.25, p = .0002, Cohen’s d = 0.20, BF₁₀ = 162.85; Fig. 3). For spoken color targets, there was no significant difference between congruent (mean RT = 392.36 ms, SD = 78.31) and incongruent conditions (mean RT = 394.81 ms, SD = 76.09; t(33) = 1.07, p = .29, Cohen’s d = 0.06, BF₁₀ = 0.31; Fig. 3). Moreover, participants responded faster recognizing auditory color (mean RT = 393.59 ms, SD = 77.20) than shape (mean RT = 409.46 ms, SD = 89.51) targets, β = 11.38, SE = 2.74, t = 4.16, p < .01.

Compare the two Experiments

In order to examine the effect of sensory input (visual vs auditory), further analysis using Experiment (Experiment 1 vs Experiment 2), congruency condition (congruent vs incongruent), and the interaction between Experiment and congreuncy condition as fixed factors, and participants and target as random factors (RT ~ Congruency condition + Experiment + Congruency condition*Experiment + (1|Target) + (1|Subject)) showed that there was no significant main effect of Experiment, β = -12.27, SE = 12.78, t = -0.96, p = .34, no significant interaction effect between Experiment and congruency condition, β = 0.69, SE = 2.84, t = 0.24, p = .81. The non-significant interaction effect between Experiment and congruency condition suggested that there was little difference between the two Experiment on the congruency effect of CSAs. Significant main effect of congruency condition on RTs was observed, β = 7.44, SE = 1.42, t = 5.26, p < .001, indicating that participants responded faster in congruent (mean RT = 389.51ms, SD = 79.56) than incongruent (mean RT = 402.45 ms, SD = 86.57) conditions. Thus, there was no significant difference on sensory input of CSAs affecting recognition efficiency for color and shape features.

For color targets across the two experiments, participants showed a moderate congruency effect of CSAs that RTs in congruent (mean RT = 387.18 ms, SD = 77.92) conditions were moderately shorter than incongruent conditions (mean RT = 391.83 ms, SD = 81.10; t(63) = 2.69, p = .009, Cohen’s d = 0.10, BF₁₀ = 3.67). For shape targets across the two experiments, participants responded significantly faster in congruent (mean RT = 397.41 ms, SD = 83.78) than incongruent conditions (mean RT = 407.49 ms, SD = 89.00; t(63) = 5.42, p < .001, Cohen’s d = 0.18, BF₁₀ = 15468.35).

In the present study, we used two Go/No-go experiments to examine the congruency priming effect of CSAs on recognition efficiency for color and shape features, in which color or shape target was presented in visual word form (Experiment 1) and spoken word form (Experiment 2). Results showed a significant congruency effect of CSAs that participants recognized color or shape faster in congruent than incongruent priming conditions in both experiments. Thus, CSAs could be strong enough to influence visual recognition of color and shape features. Moreover, the sensory input modalities (in visual words or spoken words) played little effect in the congruency priming effect of CSAs.

Previous studies showed that CSAs could be tested by indirect experimental methods (e.g., IATs; Chen et al., 2015b), that participants responded faster when congruent color-shape mapping shared the same behavioral responses than incongruent conditions (e.g., if response to color and to associated-shape should be produced using the same hand). Here, using two Go/No-go tasks, we complemented the existing literature by showing a priming effect of congruent CSAs on biasing recognition efficiency for color and shape features. Those results provided further evidence that learned CSAs could be strong to affect low-level visual processing with discrimination for color and shape features.

Moreover, the current study using an exploratory examination of the target input modality, that there was little effect on the target modality (visual and auditory word forms) influencing congruence effect of CSAs on visual discriminations, suggesting that the congruency effect of CSAs could also be tested by auditory word input. CSAs are suggested to be semantic mediated correspondences, which can be constructed through language learning experience, and requires hearing phonological information (Chen et al., 2015a; Chen et al., 2016). Chen et al. (2016) reported that deaf people showed weak congruency effect of CSAs on behavior performances tested by IATs, but revealed similar pattern of CSAs when tested by explicit matching task, compared with hearing people. The present findings provided further evidence for semantic correspondence account, that both visual and auditory word forms of color and shape affected recognition efficiency for associated color and shape features. Thus, CSAs might be high-level associations more than perceptual associations.

It could also be possible that CSAs are learned through repeated exposure of color-shape conjunctions in the environment (Statistic correspondence; Jacobsen, 2002; Hanada, 2019; Spence, 2011, 2018). For instance, the red-circle associations might be influenced by a salient cultural mediation factor; e.g., the Japanese flag has a red circle in the center of a white background. Moreover, these associations were also observed in various other countries. Albertazzi et al. (2013) observed circle-red, square-red, and triangle-yellow associations among Italian people. Kharkhurin (2012) found a square-blue association in participants from varied cultural backgrounds. Makin and Wuerger (2013) observed a marginal effect for square-red and triangle-yellow associations in British participants. Consequently, it would be helpful to investigate possible cultural differences/statistical learning effect on the implicit strength of CSAs. Furthermore, Dreksler and Spence (2019) suggested the affective correspondence hypothesis, that the liking and arousal appraisals have an effect on CSAs, such as the most (least) preferred/arousal colors and shapes tended to be associated together. It was suggested that the later-learned semantic/statistic/emotion correspondences could be interacted to mediate sensory information processing, that congruent associations accelerate behavioral responses (Spence & Deroy, 2013; Maimon et al., 2020; Wallmark et al., 2021).

It should be noted that the congruency effect of CSAs failed to affect recognition for color targets in Experiment 2 (BF10 = 0.31), suggesting that recognition for visual color features primed from spoken color words target might be less interfered with the presented visual shapes. Moreover, the effect of congruent CSAs for color targets was also moderate in Experiment 1 (BF10 = 6.93). Those results might suggest a less interference effect with congruent-shape features for color recognition than congruent-color features for shape recognition. Thus, the congruency effect of CSAs interfered more for shape recognition than for color recognition. Furthermore, recognition for colors were faster than shapes in both experiments, consist with previous findings suggesting that colors could be processed faster than shapes (Cohen, 1997; Livingstone & Hubel, 1988).

One limitation of the present study is that the experiment design with “Go/No-go” task mainly used to compare the response times elicited by congruent and incongruent colored-shape stimuli in “Go” trials, while the response accuracy of “No-go” trials was not included. Future studies may adopt more sensitive experiment design with “Go/No-go” trials to examine both response times and accuracy to test the strength of CSAs. For instance, the trials are set at a rapid pace, so that responses have to be made quickly, and more likely to get false alarms (Meule, 2017; Wessel, 2018). We used 1500 ms here, which was on the long wait for participants to get ready for the next trial, limited the likelihood of making false alarms. Future studies may use signal detection analysis with hits vs. false alarms across conditions to test the congruency effect of CSAs. We hypothesis that participants make more errors for congruent conditions in “No-go” trails, to show that CSAs may interfere with inhibitory control and motor activity.

In summary, we demonstrated congruency priming effect of color–shape associations on recognition efficiency for color and shape features, irrespective of target input modality (visual or auditory). Those results provided further evidence that learned high-level semantic associations could be strong to affect low-level visual feature discrimination. color–shape associations can be applicable in visual product design with colors and shapes, that congruent color and shape features might be processed more readily and easily. Future studies may address the congruency effect of color–shape associations on visual processes, such as visual search and visual working memory, to reveal the nature of those within-visual color–shape associations.

Supplementary material

Supplemental materials, the raw data, and R code for this article is available online. https://osf.io/g78nv/

Fundings and acknowledgments

This work was supported by Grant-in-Aid for Scientific Research (17F17008,17H00753, 20K22296, 21K13759, and 22H00090) from Japan Society for the Promotion of Science. We would like to thank the anonymous reviewers for their suggestions and comments.

Competing interests

The authors have no conflict of interest directly relevant to the content of this article.

Albertazzi, L., Da Pos, O., Canal, L., Micciolo, R., Malfatti, M., & Vescovi, M. (2013). The hue of shapes, J. Exp. Psychol. Hum. Percept. Perform. 39, 37–47.

Albertazzi, L., Malfatti, M., Canal, L., & Micciolo, R. (2015). The hue of angles—Was Kandinsky right?. Art. Percep. 3(1), 81–92.

Bates, D., Maechler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67(1), 1–48. http://dx.doi.org/10. 18637/jss.v067.i01

Chen, N., Tanaka, K. & Watanabe, K. (2015a). Color-shape associations revealed with implicit association tests, Plos ONE. 10(1), e0116954.

Chen, N., Tanaka, K., Matsuyoshi, D. & Watanabe, K. (2015b). Associations between color and shape in Japanese observers, Psychol. Aesthet. Creat. Arts. 9, 101–110.

Chen, N., Tanaka, K., Namatame, M. & Watanabe, K. (2016). Color-shape associations in deaf and hearing People, Front. Psychol. 7, 355.

Cohen, D. J. (1997). Visual detection and perceptual independence: Assessing color and form, Percept. Psychophys. 59(4), 623–635.

Dreksler, N., & Spence, C. (2019). A critical analysis of colour-shape correspondences: Examining the replicability of colour-shape Associations. I-Perception. 10, 2041669519834042.

Dumitrescu, A. (2011). New researches regarding relationship between elementary geometric shapes and basic colors, in: Annals of DAAAM & Proceedings, pp.1041-1043.

Greenwald, A. G., McGhee, D. E. & Schwartz, J. L. (1998). Measuring individual differences in implicit cognition: the implicit association test, J. Pers. Soc. Psychol. 74, 1464–1480.

Hanada, M. (2019). Associations of visual forms with colors: The minor role of emotion as the mediator. Color Res. Appl. 44, 568–580.

Holmes, T. & Zanker, J. M. (2013). Investigating preferences for color-shape combinations with gaze driven optimization method based on evolutionary algorithms, Front Psychol. 4, 926.

Jacobsen, T. (2002). Kandinsky's questionnaire revisited: fundamental correspondence of basic colors and forms?. Percept. Mot. Ski. 95, 903–913.

Jacobsen, T., & Wolsdorff, C. (2007). Does history affect aesthetic preference? Kandinsky's teaching of colour-form correspondence, empirical aesthetics, and the Bauhaus. Des. J. 10, 16–27.

Kandinsky, W. (1912/1994). On the spiritual in art. New York: Da Capo Press.

Kandinsky, W. (1947). Point and line to plane. New York: Solomon R. Guggenheim Foundation, First German edition.

Kharkhurin, A.V. (2012). Is triangle really yellow? An empirical investigation of Kandinsky’s correspondence theory, Emp. Stud. Arts. 30, 167–182.

Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. (2017). lmerTest package: tests in linear mixed effects models. J. Stat. Softw. 82, 1–26.

Livingstone, M. & Hubel, D. (1988). Segregation of form, color, movement, and depth: anatomy, physiology, and perception, Science. 240(4853), 740–749.

Lupton, E. & Miller, J. A. (Eds.). (1993). Dreieck, quadrat, kreis: Bauhaus und design-theorie heute [Triangle, square, circle: Bauhaus and design theory theory today]. Basel, Switzerland: Birkhauser.

Makin, A. & Wuerger, S. (2013). The IAT shows no evidence for Kandinsky's color-shape associations, Front. Psychol. 4, 616.

Malfatti, M. (2014). Shape-to-color associations in non-synesthetes: Perceptual, emotional, and cognitive aspects (Doctoral dissertation). University of Trento, Italy. Retrieved from http://eprints-phd.biblio. unitn.it/1336/

Meule, A. (2017). Reporting and interpreting task performance in go/no-go affective shifting tasks. Frontiers in Psychology, 8, 701.

Morey, R. D. & Rouder, J. N. (2018). BayesFactor: Computation of Bayes Factors for Common Designs. R package version 0.9.12–4.2.

Nosek, B. A. & Banaji, M. R. (2001). The go/no-go association task, Soc. Cogn. 19, 625–666.

Schulz, K. P., Fan, J., Magidina, O., Marks, D. J., Hahn, B. & Halperin, J. M. (2007). Does the emotional go/no-go task really measure behavioral inhibition? Convergence with measures on a non-emotional analog. Arch. Clin. Neuropsych. 22, 151–160.

Spence, C. (2011). Crossmodal correspondences: A tutorial review, Attent. Percept. Psychophys. 73, 971–995.

Spence, C. (2018). Crossmodal correspondences: A tutorial review. In D. Howes (Ed.), Senses and sensation: Critical and primary sources (Vol. III, pp. 91–125). London, England: Bloomsbury Academic.

Spence, C. & Deroy, O. (2013). How automatic are crossmodal correspondences? Conscious. Cogn. 22, 245–260. doi: 10.1016/j.concog.2012.12.006

Velasco, C., Wan, X., Knoeferle, K., Zhou, X., Salgado-Montejo, A., & Spence, C. (2015). Searching for flavor labels in food products: the influence of color-flavor congruence and association strength. Front. Psychol. 6, 301.

Wessel, J. R. (2018). Prepotent motor activity and inhibitory control demands in different variants of the go/no‐go paradigm. Psychophysiology, 55(3), e12871.

Download PDF

Version 1

posted

You are reading this latest preprint version

Effect of Color–Shape Associations on visual feature discrimination

Status:

Version 1

Abstract

Figures

1. Introduction

2. Experiment 1

2.1 Method

2.1.1 Participants

2.1.2 Apparatus and stimuli

2.1.3 Procedure

2.1.4 Data analysis

2.2 Results

3. Experiment 2

3.1 Method

3.1.1 Participants

3.1.2 Apparatus and Procedure

3.2 Results

4. Discussion

Declarations

References

Status:

Version 1