Shape variety of food can boost its visual appeal ☆

We investigated whether food shape and its variety within a group affect visual appeal using a four-shaped fast-food chicken product known as Chicken McNuggets ® . In Experiment 1, participants rated the visual appeal of each nugget shape both individually and when presented in groups of variously shaped nuggets. The results revealed that the rounder nugget was less visually appealing than those of other shapes. Additionally, assort-ments featuring various shapes were rated as more appealing than those of a single shape. In Experiment 2, the visual appeal of individual nuggets presented in groups and alone was assessed using a visual analog scale. The visual appeal of one specific nugget (target nugget) was higher when presented in a group than alone. Furthermore, a target nugget presented in a group with various shapes was more visually appealing than when presented in a group with the same shape, regardless of the shape of the target nugget. These results suggest that a food product with low visual appeal can be perceived as more appealing when it is presented alongside various food shapes. Indeed, the variety of food shapes presented in a group affected the perceived appeal both of the group and of the individual food item. These findings offer novel insights into the impact of food ’ s visual variation on its appeal.

Shape is one of the most crucial factors influencing the visual appeal of food.Studies in esthetics and psychology have revealed that people generally prefer round shapes rather than angles (Bar & Neta, 2006, 2007;Leder & Carbon, 2005;Westerman et al., 2012).Round shapes can be related to and generate more positive affective feelings (e.g., Bar & Neta, 2006;Blazhenkova & Kumar, 2018).Angular shapes can be related to threat (Bar & Neta, 2006, 2007) and more violent/aggressive feelings (e.g., fear: Westerman et al., 2012).Regarding food shape, hedonic or sweet foods with rounder shapes (e.g., chocolate cake) tend to be visually preferred to those with angular shapes (Zhou et al., 2021).However, it should be noted that the visual appeal of rounder utilitarian foods (e.g., cereal bread) is no different from that of more angular utilitarian foods (Zhou et al., 2021).We see, therefore, that the shape of a food can influence its visual appeal, but that the influence may depend on the food.
In addition to the individual shapes themselves, the variety of shapes included in a product or dish may also enhance its visual appeal.For example, Mielby et al. (2012) examined the relationship between the Abbreviations: SRT, Shape Rating Task; VRT, Variety Rating Task; VAS, Visual Analog Scale.☆ A part of this paper was presented at the Annual Meeting of Japanese Society of Sensory Evaluation (November 20, 2022, teleconference).
visual preference for foods on a plate and variety in the food's appearance.Mielby et al.'s participants performed incomplete rankings of their visual preferences for fruits and vegetables.These rankings were designed using eight levels of variety in factors such as cut, color, number of products, type, and combination of products.The results showed that when the variety of food appearance increases, visual preference also tends to increase; however, when variety is excessively high, visual preference decreases.Therefore, the study found an inverted U-shaped relationship between the variety of food appearance and its visual appeal.To the best of our knowledge, Mielby et al. (2012) and other previous studies (e.g., Bar & Neta, 2006;Zhou et al., 2021) did not necessarily focus solely on the shape variety of a food product for visual appeal.If the shape of food plays an important role in the consumer's perception of the variety of foods, then the same foods with a variety of shapes may be perceived as more appealing than a group of the same foods with the same shape.
Additionally, whether or not a variety of food shapes within a group biases perceptions of the visual appeal of individual food items (i.e., one of several food items on a plate) remains unknown.Here, we also wondered whether a variety of food shapes boosts the perceived visual appeal of an individual food item within the group.One cue for understanding the effect of variety on food preferences may be found in the field of cognitive psychology.Walker and Vul (2014) found that an individual's face when seen in a group was perceived as more attractive than the same individual's face alone (however, see studies that replicated these results: Carragher, Lawrence, Thomas, & Nicholls, 2018; and those that failed to replicate them: Ojiro et al., 2015;van Osch, Blanken, Meijs, & van Wolferen, 2015).This phenomenon is known as the cheerleader effect.Walker and Vul (2014) explained it as follows: First, when multiple faces are presented, an individual's visual system computes the average of the faces (Ariely, 2001;Chong & Treisman, 2003).Second, the participant's perception of each face in the group is biased toward the computed average (e.g., Brady & Alvarez, 2011).Finally, an individual's face is rated as more attractive within the group than alone because people feel attracted to faces that are closer to the average (Langlois & Roggman, 1990).The cheerleader effect implies that information surrounding the target stimulus is used as a cue or criterion for rating individual targets.
Regarding food rating, an individual food item in a group may be perceived as more appealing than the individual item presented alone.However, if the cheerleader effect is observed with food, it would not be attributed to the mechanism demonstrated by face averaging.The cheerleader effect for food would be due to multiple factors related to the perception of food.First, several food items arranged and presented together tend to be preferred over a single item presented alone (Woods, Michel, & Spence, 2016).Second, as we predicted above, several food shapes presented together may increase the visual appeal of the food overall because of the variation in its appearance (Mielby et al., 2012).Third, people may rate the visual appeal of a target food using information surrounding the target, such as the overall visual appeal.For example, the color of the plate under the target food significantly influences the visual appeal of the target food (e.g., Piqueras-Fiszman, Giboreau, & Spence, 2013).Taking all of this together, we assumed the following: Arranging several food items together and presenting a variety of food shapes boosts the overall food visual appeal to a higher level.This boost in visual appeal is used as a cue for rating individual food items, and may result in a positive bias towards the visual appeal of individual food items.
Using a food product that allows us to keep all factors besides shape constant, we are able to put our hypothesis to the test within the context of a controlled study.Y. Nishida et al. first executive chef, who developed the nugget from an idea based on tempura (The Blade, April 4, 1985).As of December 2021, when the present study was conducted, the nuggets contained chicken thigh meat, breast meat, and skin.The recipe was modified in February 2023, and, since then, the composition of the nuggets has been chicken breast meat and skin.According to McDonald's, the nuggets' mass, size, and frying time are almost uniform and they are randomly placed in boxes when served.Thus, nuggets are an ideal stimulus for manipulating the visual variety of food because they are identical except in shape and each of the four shapes are extremely consistent.Furthermore, Japanese participants are familiar with nuggets.Some individuals may have a degree of food neophobia (e.g., Dovey, Staples, Gibson, & Halford, 2008 for a review), a personality trait in which a person tends to reject novel and unknown foods.The use of unfamiliar foods in an experiment may make it difficult for such individuals to rate those foods; therefore, participants may find it easier to rate the visual appeal of nuggets thanks to their familiarity.
Thus, here we elucidated the influence of shape variations of the same food product on visual appeal, using nuggets as stimuli in two psychological experiments.While rounded hedonic foods tend to be preferred (Zhou et al., 2021), it remains unknown if this also holds true for nuggets.Therefore, in Experiment 1, we assessed participants' feelings for the visual appeal of each individual nugget shape and also tested the effects of shape variety within a group on visual appeal.We hypothesized that the greater the variety of shapes among the nuggets within a group, the more visually appealing they would be (i.e., the shape variety effect).If this prediction is true, the nuggets will become more visually appealing as they are presented together with an increasing number of shapes.We did not anticipate a decrease in the visual appeal of the nuggets due to an excessive variety of food appearance because there were only four different nugget shapes in our study.In other words, the present study did not aim to replicate the inverted U-shaped relationship between visual preference and the appearance of food variety described by Mielby et al. (2012).
In Experiment 2, we hypothesized that an individual nugget within a nugget group would be rated as more appealing than a single nugget alone (i.e., the nugget cheerleader effect) and examined whether the variety of surrounding shapes would make any target shape more appealing.

Experiment 1
First, this experiment was designed to examine whether a nugget's visual appeal differed according to its shape by comparing the visual appeal ratings of the four shapes: Ball, bell, bone, and boot.To test this, participants performed the Shape Rating Task (SRT).Second, it was used to investigate the effects of a group of nuggets of various shapes on visual appeal using the Variety Rating Task (VRT).Indeed, we focused on examining how different varieties of shapes influence perceived visual appeal based on relative ratings.Therefore, participants rated the visual appeal of a single nugget shape and of a collection of nuggets with various shapes compared to other stimuli during shape rating and variety rating tasks.

Participants and design
We designed Experiment 1 as a one-factor, within-participant experiment.Initially, we aimed to produce as much statistical power as possible under the limitations imposed on research resources due to the COVID-19 pandemic.For example, in a previous study of face attractiveness, Walker and Vul (2014), Experiments 1, 2, and 3 had 25, 18, and 20 participants, respectively, and they detected differences among experimental conditions.Given the constraints on resources, previous studies' sample sizes, and our own prior experience, we concluded that having approximately 20 participants would enable us to maximize statistical power.A total of 24 people (17 females, 7 males; M age = 21.4 years, SD age = 0.57) volunteered to participate in the experiment.Due to the COVID-19 restrictions that were in effect during the study period, we took several precautions to ensure the safety and well-being of participants and experimenters: Before the experiment, all participants and researchers washed and sanitized their hands, had their temperatures taken, and wore surgical masks.We consistently implemented these precautions across all experiments to minimize risks associated with the pandemic.

Materials and apparatus
In the SRT, photographs of each nugget shape were used as visual stimuli.Stimuli were photographed using an IRIS visual analyzer (Alpha Mos, Japan).Calibration was performed using a color checker prior to photography.The original photographs were 1871 × 1871 pixels in size and were adjusted to a size of 91 × 91 pixels for display in Microsoft PowerPoint (Fig. 1; see also Supplementary Fig. 1).The adjusted stimuli were presented on a 14-inch monitor (Full HD 1920 × 1080) on a personal computer (Acer, SF114-32-A14U/S).To confirm the relationship between the visual appeal of nugget shapes and the physical characteristics of each nugget's shape, we took supplemental measurements of the roundness and aspect ratio of each of the nugget shapes using a visual analyzer.
In the VRT, the stimuli consisted of 15 paper plates containing five nuggets each.The paper plates had a diameter of 18 cm.Since McDonald's nuggets sold in the Japanese market come in packs of five, we arranged five nuggets on each plate such that they did not physically overlap.Each plate contained a different combination of shapes: Four plates of one shape (i.e., all five nuggets were the same shape), six plates of two shapes, four plates of three shapes, and one plate of four shapes (Table 1).While there are more than 50 possible combinations of nugget shapes, it would be difficult for participants to rate all combinations relative to each other.Therefore, we presented participants with 15 combinations, which we considered to be a moderate amount, neither too many nor too few.These 15 combinations were chosen so that all shapes were presented in equal or close to equal numbers (i.e., 20 ball shapes, 17 bell shapes, 19 boot shapes, and 19 bone shapes).For combinations where two or more shapes were mixed, the number of nuggets of each shape was as equal as possible.A 143 × 100 cm sheet of graph paper was fixed to a table with tape to prevent the paper from moving."Does not look tasty" was marked on the left side of the graph paper, and "Looks tasty" on the right side.Fifteen plates of nuggets were placed together on graph paper (Fig. 2A).

Procedure
The participants were tested individually.They performed the SRT first.Four nugget stimuli of different shapes were displayed as

Table 1
Combination of nugget shape types in Experiment 1.

Plate name
Number of shape types Type of shape Note.A, B, C, and D represent ball, bell, boot, and bone, respectively.
Y. Nishida et al.PowerPoint slides on a PC monitor.The stimuli were placed in horizontal rows at random positions (Supplementary Fig. 1).Participants placed less tasty-looking stimuli to the left side of the frame on the slide and more tasty-looking stimuli to the right, without a time limit.The participants were instructed that the vertical orientation was not relevant to their rating.Following the SRT, the participants performed the VRT.They were asked to stand before a desk where 15 plates with nuggets were placed.The initial positions of the nugget plates were randomized.Subsequently, they rated the plates by freely rearranging them horizontally from less tasty-looking (towards the left) to tasty-looking (towards the right).The participants were instructed that the vertical orientation was not relevant to their ratings.After the rearrangement, participants responded to a Google Form questionnaire asking about their age, sex, frequency of nugget consumption, and preferred shape.There was no time limit for responding.

Data analysis
Roundness is a geometric parameter describing the deviation of a stimulus from a perfectly circular shape, indicating how close the object is to circularity.The closer the object is to a circle, the greater the roundness value.We refer to aspect ratio as the proportional relationship between the width and height of a nugget, calculated as the ratio of an object modeled as an ellipse's major diameter to its minor diameter.The larger this value is, the closer the object's shape is to a circle.Each value was multiplied by 100, after which they ranged from 0 to 100.The equations for obtaining each value are provided in the Supplementary Equations 1 and 2.
After the experiment, we conducted power analyses using the same approach as an a priori power analysis by G*Power (Version 3.1.9.4).We assumed a one-factor within-participant analysis of variance (ANOVA) with 4 levels (SRT: nugget shape; VRT: number of different shapes).The results indicated that, for Experiment 1, a sample size of 24 participants would provide adequate power (80%) to detect a medium effect size (f = 0.25) of main effect at a significance level of 5% (also see Supplementary Fig. 5).In accordance with Heckman, Davis, and Crowson (2022), we analyzed 95% confidence intervals (CIs) for effect size as an alternative to conducting post hoc power calculations.Ultimately, we reported 95% CIs for both experiments.
For the SRT analysis, we defined the degree of visual appeal as the distance (pixels) between the image stimulus and the left edge of the frame as zero.That is, if the distance was greater than zero, the participants rated the nuggets as more tasty-looking.We corrected each value to a percentage of the frame length.For the VRT analysis, we recorded the horizontal positions of each plate placed side-by-side on the graph paper.The distance to each plate was measured starting from the left end of the graph and converted into percentages.This value was defined as visual appeal for the VRT.We calculated the mean of visual appeal for each number of shape types and participants.
Data from three participants were excluded because of technical problems.Twenty-one participants were included in the analysis.All of the reported statistics are available in the Supplemental Material.We conducted a one-way repeated measures ANOVA for the SRT and VRT.In the analysis of the VRT, we performed a supplementary analysis for "with-ball" and "without-ball" conditions to test the effect of a variety of shapes for each.In other words, the data were divided based on whether the plates included or did not include ball-shaped nuggets.Because the plates of all four shapes (4ABCD) included the ball-shaped nugget, the maximum number of shape types in the with-ball condition was 4. In contrast, in the without-ball condition, the maximum was 3. Since doing two-way repeated measures ANOVA without an equal number of levels on the same factor is not feasible, we separated the data as "with-ball" and "without-ball" and performed one-way repeated measures ANOVA for each.
All statistical analyses in this study were conducted using the R software (version 4.1.1;R Core Team, 2021).We calculated the effect size of η p 2 , Cohen's d, and the 95% CIs using the 'MOTE' and 'rstatix' packages in R. For Cohen's d, around 0.2 is commonly considered small, 0.5 is medium, and 0.8 is large (Cohen, 1988).We applied the Holm-Bonferroni correction for multiple comparisons.Adjusted p-values were provided only for statistically significant results.Meanwhile, original p-values in multiple comparisons were noted in Supplementary Tables 1 and 2 The significance level was 5%.
Because the analysis of the SRT revealed that the ball-shaped nugget had the lowest visual appeal and that there was no significant difference in visual appeal among the other shapes, the ball-shaped nugget could have the most decisive influence on the visual appeal of the plates of nuggets during the VRT.As Table 1 shows, the more types of nugget shapes on the plates, the smaller the number of ball-shaped nuggets included.At this point in time, it was unclear whether increased types of nugget shapes and fewer ball-shaped nuggets would increase visual appeal or whether only one factor (i.e., either increased types of shapes or fewer ball-shaped nuggets) would enhance visual appeal.Thus, as independent datasets, we conducted separate one-way repeated measures ANOVAs to examine the influence of the number of shape types on visual appeal in both the with-ball and without-ball conditions (Fig. 3).One-way repeated measures ANOVAs revealed a significant main effect of the number of types of shapes in both the with-ball and without-ball conditions: F( 3 However, it should be noted that the effect size was larger in the with-ball condition (see also Supplementary Table 2).

Discussion
Regarding the SRT, we found that the ball shape was the least appealing.As shown in Table 2, the ball shape was confirmed to be the Note.The roundness value indicates the degree of circularity of an object.Higher values suggest a closer resemblance to a circle.The aspect ratio value represents the ratio of an object's height to its width: The ratio of the major to the minor diameter of the object is modeled as an ellipse.Larger values indicate a shape closer to a circle.Both roundness and aspect ratio values fall between 0 and 100.Y. Nishida et al. closest to a circle among the four shapes.It was unexpected that a rounder nugget would be perceived as the least appealing.The reason for these findings is highlighted in the General discussion section below.
Regarding the VRT, our results indicate that nugget plates with more than one shape are appealing, reflecting the shape variety effect.Mielby et al. (2012) showed that a variety of food appearances presented on a single plate increases visual preference for the food.The participants in our study may have perceived food variety due to the variety of shapes, which, in turn, may have increased their appetite and preferences for the nuggets.Alternatively, the variety of shapes could reflect the normalcy (nuggets as participants are used to seeing in daily life) of the food's appearance, which, in turn, led to a relatively positive evaluation.However, more evidence is needed to determine why a variety of shapes promotes visual appeal.For now, at least, our hypothesis that the presentation of a variety of shapes increases the visual appeal of food was supported.
It is also important to note that there was an exception in the supplemental analysis.As seen in Table 1, in most cases, the number of ballshaped nuggets included on a plate decreased as the number of nugget shape types increased: There were five ball-shaped nuggets for oneshape plates, three for two-shape plates, two for three-shaped plates, and one for the four-shape plate.The only exception was the plate labelled 3ACD, which had one ball-shaped nugget.There was a possibility that the exception plate may have affected the results of the withball condition.If this possibility were true, the visual appeal of the nuggets on the exception plate would differ from that of the nuggets on plates 3ABC and 3ABD.However, there was no noticeable difference (see Supplementary Fig. 3).The differences in the shape variety effect for with-ball and without-ball conditions are discussed further in the General discussion section.

Experiment 2
In Experiment 1, we revealed that the presentation of a variety of nugget shapes enhanced the visual appeal of the group of five nuggets on a plate.The first question for Experiment 2 was whether an individual nugget within a nugget group is rated as more appealing than the same individual nugget alone (i.e., the nugget cheerleader effect).A further question was whether or not variety in the nugget shapes within a group would enhance an individual nugget's visual appeal.To answer the second question, we explored whether the strength of the nugget cheerleader effect differs depending on the shape of an individual target nugget.Therefore, the nugget groups were divided into homogeneous (the surrounding nuggets are all the same) and heterogeneous (the shapes of the surrounding nuggets are all different) groups.An absolute rating of the visual appeal of nugget shapes was conducted to confirm consistency with the relative rating results from Experiment 1.

Participants and design
This experiment adopted a within-participant design with two factors: Presentation type (group and single) and element nugget shape (ball, bell, bone, boot, and all).We initially considered approximately 20 participants based on previous experience and the number of participants in Walker and Vul's (2014) study to ensure adequate statistical power.A total of 22 people (female = 15, male = 7, M age = 21.5 years, SD age = 2.02) participated in Experiment 2.

Materials and apparatus
The visual food stimuli were photographed using an IRIS visual analyzer with four nuggets arranged on a plate in two rows and two columns (Fig. 4A).For the group condition, there were five images (1470 × 1470 pixels) of nugget groups: Four images each displaying groupings of the same shape (ball, bell, bone, and boot conditions, respectively) and one displaying all four nugget shapes (all-type condition).For the single condition, we cropped each nugget group image to produce single-nugget images (486 × 486 pixels).There was a total of 20 single-condition images.The size of the stimuli presented on the screen was adjusted such that the size of the nuggets was the same in both the group and single conditions.

Procedure
A maximum of three participants participated in the experiment during any one session.Because the procedures in Experiment 2 were complicated, participants completed eight practice trials ensure their understanding of the procedure.The stimuli images for the practice trials were groups of chocolates (four trials) and a single chocolate (four trials).The chocolate shapes were round or square.For the actual experiment, there were 20 trials for the single condition and 20 trials for the group condition, and single and group condition stimuli were presented randomly.The flow of each trial is shown in Fig. 4B and C. For the single condition, after a fixation point (500 ms), a single nugget image was presented for 2000 ms (Fig. 4C).The image then disappeared, a visual analog scale (VAS) appeared on the display, and participants rated how tasty the nugget looked on a scale of 0-100 (0 = not tastylooking, 50 = neither tasty-looking nor not tasty-looking, 100 = tastylooking).The participants then moved on to the next trial.For the group condition, after a fixation point of 500 ms, an image of a group of nuggets was displayed for 1000 ms (Fig. 4B).Subsequently, an arrow pointing to a nugget in the image was presented for 1000 ms.The image and arrow then disappeared, and the VAS appeared on the display.The participants rated how tasty-looking the nugget pointed to by the arrow looked.The order in which the images were presented and which nugget in the group was pointed to by the arrow were randomized.

Data analysis
After the experiment, we conducted a power analysis using the same approach as an a priori power analysis by G*Power (Version 3.1.9.4).We assumed a two-factor within-participant analysis of variance with two levels (presentation type: single and group) and five levels (element nugget shape: ball, bell, bone, boot, and all).The post-experiment power analysis, which is the same approach as an a priori power analysis, indicated that a sample size of 22 participants would provide adequate power (80%) to detect a medium effect size (f = 0.25) for interaction at a significance level of 5% (see also Supplementary Fig. 6).
Data from all 22 participants were included in the analyses.All of the reported statistics are available in the Supplemental Material.We performed two analyses to assess the nugget cheerleader effect.The first was conducted using a two-factor within-participant repeated measures ANOVA with two levels (presentation type: single and group) and five levels (element nugget shape: ball, bell, bone, boot, and all).For the second analysis, we divided the group condition into two subconditions: heterogeneous and homogeneous.We conducted a twofactor within-participant repeated measures ANOVA with three levels (presentation type: heterogeneous, homogeneous, and single) and four levels (target nugget shape: ball, bell, bone, and boot).Note, however, that because the latter analysis was not pre-designed, the number of datasets for each participant for each presentation type was 1 for heterogeneous, 4 for homogeneous, and 5 for single, which are not equal.These analyses were conducted by averaging the visual appeal ratings for each condition and participant.
All statistical analyses in this study were conducted using the R software (version 4.1.1;R Core Team, 2021).We calculated the effect size of η p 2 , Cohen's d, and these 95% CIs using 'MOTE' and 'rstatix' packages in R. A simple effects test was performed when the interaction was significant.We applied the Holm-Bonferroni correction for multiple comparisons.Adjusted p-values were provided only for statistically significant results.All original p-values in multiple comparisons were noted in Supplementary Tables 4, 5, and 6.The significance level was 5%.

Second analysis: presentation type and target nugget shape
Fig. 6 shows the average visual appeal for the presentation and target nugget shape types.The ANOVA revealed a significant main effect of presentation type, F( 2 Y. Nishida et al. nuggets were rated as less visually appealing than other shapes: adjusted ps < 0.05, Cohen's ds > 0.65.No significant differences were found between other shape combinations.The interaction effect was nonsignificant: F(6, 126) = 0.57, p = 0.754, η p 2 = 0.026, 95% CI [0.000, 0.052].
We have included more detailed analysis results in Supplemental Tables 3, 4, 5, and 6.

Discussion
The results revealed a nugget cheerleader effect from the perspective of element nugget shape type and target nugget type.The results of the first analysis showed that the visual appeal of individual nuggets in the group condition was higher than that in the single condition, regardless of the element nugget shape type.Our hypothesis that a pronounced nugget cheerleader effect would be observed in the all-shape-types condition was not strongly supported because the interaction effect was nonsignificant.In contrast, the second analysis showed that the visual appeal of the target nuggets was greater in the heterogeneous condition than in the single condition, but was similar between the homogeneous and single conditions.These results suggest that the nugget cheerleader effect is less likely to occur when the shape of the surrounding nuggets is the same as that of the target nugget.As a moderate variety in the appearance of food products has been shown to increase general preference (e.g., Mielby et al., 2012), our participants may have perceived moderate visual variety in the heterogeneous nugget group, leading to a preference for an individual nugget within heterogeneous groups.
Considering the varying amount of data available among presentation types (heterogeneous, homogeneous, and single), it is essential to exercise caution when evaluating the reliability of the second analysis.However, at the very least, we can observe that the nugget cheerleader effect is more likely to be robust when various shapes are included in a group.Nevertheless, it should be noted that we did not find a significant difference between homogeneous and heterogeneous conditions.This will be discussed in more detail in the General discussion.In addition, the results of Experiment 2 reflected those of Experiment 1, revealing less visual appeal for the ball shape than for other shapes in absolute as well as relative ratings.

General discussion
This study investigated the impact of food shape and its variety within a group on visual appeal.The results of the SRT in Experiment 1 and the second analysis in Experiment 2 showed that the visual appeal of the ball-shaped nugget was less than that of the other shapes.For the VRT in Experiment 1, visual appeal increased with the variety of nugget shapes presented at once (the shape variety effect).The results of Experiment 2 revealed that the visual appeal of individual nuggets was higher when they were presented in a group than when presented alone (the nugget cheerleader effect).Subsequent analysis indicated that the visual appeal of individual nuggets was greater in the heterogeneous condition than in the single condition but not in the homogeneous condition.These findings align with related research by Mielby et al. (2012), suggesting a potential link between preference and the visual variety of a food.In their study, Mielby et al. manipulated the appearance of food, examining the impact of visual variety using foods of different shapes, colors, and ingredients.In our experiments, the color and ingredients of the nuggets were left unchanged, and only the way shapes were presented was manipulated to examine the effect of shape variety.Our investigation yielded novel insights, indicating that the presentation of more food shapes enhances perceptions of the visual appeal of the food.
Here, we will focus on and discuss three possible reasons for the shape variety effect found in Experiment 1: First, the presentation of multiple nugget shapes was considered a moderate visual variety pattern, and variety led to increased preference for nuggets.Mielby et al. (2012) found an inverted U-shaped relationship between visual preference and perceived complexity, revealing that food preference peaked when food on a plate was presented with moderate visual variation, neither too little nor too much.Mielby et al.'s investigation is rooted in Berlyne's (1970) collative motivation model.According to Berlyne's model, low-complexity visual patterns are not stimulating and leave observers indifferent.Highly complex patterns are difficult to grasp and experience.Thus, patterns with medium (or optimum) complexity levels are most preferred.If Mielby et al.'s findings and Berlyne's model are correct, the visual appeal of food is reduced if it has a too great a variety of shapes.If we were to test this, we would need a huge number of food items to simultaneously present to participants.At the same time, however, the number of food items would imply greater portions, affecting the visual appeal.Some studies have suggested a reduced visual appeal of foods with oversized portions (e.g., Cornil & Chandon, 2016;Toepel et al., 2015).To further clarify the inverted U-shaped relationship between shape variety and visual appeal, it may be necessary to consider the number of food items or the food quantity.
Second, the sight of the same food items with the same shapes on one plate (i.e., nuggets of the same shape on one plate) may highlight the nuggets' abnormalities, leading to a decrease in appeal.The results from Experiment 1 show that nuggets of the same shape served on the same plate received the lowest visual appeal ratings.Indeed, people occasionally avoid purchasing and consuming foods with odd shapes (e.g., De Hooge et al., 2017;Loebnitz & Grunert, 2015).This avoidance could be considered risk-averse behavior due to a perceived abnormality in the food leading to the assumption that the taste may differ from the usual (Cooremans & Geuens, 2019).For the nuggets, participants might have perceived nuggets of the same, or homogeneous, shape in a row as strange or not standard, leading to a decrease in their appeal (e.g., Loebnitz & Grunert, 2015).Alternatively, participants might have rated the visual appeal of the nuggets presented as a variety of shapes more highly because they perceived the nuggets as handmade and, therefore, as more natural (e.g., Hayward, 2014;Spence, 2017).The relationship between food shapes with less variety and an abnormal appearance is worth consideration.
Finally, the lower appeal of the ball-shaped nugget, having been rated with the lowest visual appeal, may contribute to the shape variety effect.In Experiment 1, the visual appeal of the nuggets improved as the Y. Nishida et al. number of nugget shape types increased.However, this trend may be attributed to a reduction in the frequency of the ball-shaped nuggets as the number of nugget shape types increased (Table 1; Fig. 3).In the VRT of Experiment 1, as the number of nugget shape types decreased, the number of ball-shaped nuggets on a plate increased.For example, one of the plates with a single shape held five ball-shaped nuggets (i.e., 1A in Table 1).Conversely, as the presented nugget shape types increased overall, the number of ball-shaped nuggets presented on a plate decreased; the plate with four shapes had only one ball-shaped nugget (i.e., 4ABCD in Table 1).An additional analysis of the VRT of Experiment 1 showed that in the with-ball conditions, when there were more shape types (i.e., fewer balls), the nuggets were rated as more visually appealing.We found a similar significance in the without-ball condition, but the effect size was smaller than that in the with-ball condition (see section 2.2.2 Variety rating task).These results suggest that, in this case, the ball shape is intensely involved in the shape variety effect: Opportunities to see the ball shape decreased or the ball's presence was obscured by the presence of other shapes, making the ball less noticeable.Consequently, the shape variety effect was observed.
In this study, it was not easy to determine whether it was the number of ball-shaped nuggets or the variety of shapes that affected appeal.Alternatively, both explanations could apply.At this stage, the shape variety effect should be interpreted to include both mechanisms of increased visual variety and decreased opportunities for seeing ballshaped nuggets.To clarify whether it is the influence of the ball shape itself or of the variety of shapes presented, one could conduct experiments that manipulate not only the with and without the ball shape factors but also the with and without other shapes factors to more precisely investigate whether shape variety increases a food's appeal.The shape and number of nuggets would need to be more tightly controlled.
The results of Experiment 2 revealed the nugget cheerleader effect and suggested that the variety of shapes within a group impacts the visual appeal of individual food items.The nugget cheerleader effect was observed for the heterogeneous condition but not for the homogeneous condition.Despite this, we did not find other significant differences between the heterogeneous and homogeneous conditions.The mean visual appeal was greater for single, homogeneous, and heterogeneous conditions, in that order (Fig. 6).These results have two implications.First, the nugget cheerleader effect did not occur merely because there were more nuggets.Second, only the incorporated factors of shape variety and the arrangement of several food items could have led to the increased visual appeal of an individual nugget.
The nugget cheerleader effect implies that people rate the visual appeal of individual foods within a group using the visual appeal of the food group as a whole.The presentation of a variety of nugget shapes increased the visual appeal of the nugget group, and the increased visual appeal may have been used by participants as a criterion for rating individual nuggets.As a result, a variety of shapes enhanced the visual appeal of individual nuggets.However, it is possible that these results depend on the nugget shape and that they have limitations.For example, the average appeal ratings in the heterogeneous and homogeneous conditions seemed no different for bell-shaped nuggets although the interaction effect was not significant (see section 3.2.2Second analysis: Presentation type and target nugget shape).Future research should implement more food shapes to investigate both the detailed process of the nugget cheerleader effect and the reason the nugget cheerleader effect differs depending on the target nugget (e.g., symmetry of shape or number of elements).
The results of Experiments 1 and 2 both consistently revealed a lower visual appeal for the ball-shaped nuggets, which is inconsistent with previous studies showing that round-shaped objects tend to evoke positive emotions (Bar & Neta, 2006;Blazhenkova & Kumar, 2018).People prefer round objects over angular shapes (Bar & Neta, 2006, 2007;Leder & Carbon, 2005;Westerman et al., 2012).This preference for round shapes also extends to certain foods, with rounder sweet foods more visually appealing than angular ones, but not with rounder utilitarian food (e.g., cereal bread) (Zhou et al., 2021).Therefore, the results of the present study are unlikely to be explained by positive emotions due to the roundness of the shape.Instead, as discussed below, the taste expected from the shape and the food may affect the visual appeal of the food.
The lower visual appeal of the ball-shaped nugget could, at least in part, be explained by shape-taste crossmodal correspondence (see Spence, 2011Spence, , 2023 for reviews) for reviews).As many studies have shown, rounded shapes can be associated with and emphasize the expectation of sweetness (Ngo et al., 2013;Ngo, Misra, & Spence, 2011;Ngo & Spence, 2011;Salgado-Montejo et al., 2015;Velasco, Woods, Deroy, & Spence, 2015;Wang et al., 2017).Other shapes have different associations, for example angular shapes may be related to bitter, salty, sour, or spicy (Blazhenkova & Kumar, 2018;Velasco et al., 2015).For the present study, when seeing a ball-shaped nugget, the participants might have associated the nugget with a sweet flavor.
However, they may know that a nugget's typical taste without sauce is not sweet because, as was shown by the post-event questionnaire (see Supplementary Table 7) for Experiment 1, most of the participants had experienced eating nuggets.The discrepancy between the expectation of sweetness evoked by the shape and their knowledge of the taste based on experiences may have caused dissonance in the participants, resulting in a lower appeal for the ball-shaped nuggets.In any case, whether the less visually appealing round nuggets are related to shape-taste crossmodal correspondences is worth examining.For example, future research could present participants with different shapes of nuggets and ask them what kind of taste they associate with each nugget.
Despite these significant findings, our study has several limitations to consider when interpreting the results.First, the study focused only on nuggets, and the results are not immediately generalizable to other foods.It would be worth examining diverse foods to generalize our findings.Second, the results of this study are based only in the Japanese population.The shape-taste crossmodal correspondence may differ across cultures (e.g., Bremner et al., 2013;Salgado-Montejo et al., 2015;Wan et al., 2014).The relationship between the shape of a nugget and its visual appeal, or between different shapes and their collective visual appeal, may also vary across cultures.Third, increasing the variety of food appearance may also boost the desire for foods (e.g., Rolls, Rowe, & Rolls, 1982a; see also Piqueras-Fiszman & Spence, 2014 for a review).Regarding sensory-specific satiety (Rolls, Rolls, Rowe, & Sweeney, 1981), which is the decrease in the desire to repeatedly eat the same food, Rolls, Rowe, and Rolls (1982b) demonstrated that people who were offered three types of pasta shapes ate more than those offered one shape.According to Rolls et al. (1982b), the shape of the pasta was associated with its appearance and texture; however, the influence of its appearance alone remains unclear.A paradigm that measures sensory-specific satiety and that could control for participants' hunger levels and personal preferences, could also be effective in uncovering the mechanism of the shape variety effect in detail.Fourth, in Experiments 1 and 2, no power analysis was performed prior to the experiments.In general, conducting a sample size analysis before experimenting is desirable; thus, although power analyses conducted after the experiments confirmed a sufficient sample size, the results of this study need to be replicated with sufficient planning using a preregistration system (e.g., https://osf.io/registries).Alternatively, it would be worth conducting multiple large-scale studies with diverse populations and conditions to validate the current findings, as in Woods et al. (2016) (Ns = 100-1816), a food preference study conceptually similar to the present study.

Conclusion
Increasing the variety of nugget shapes within a group could enhance the visual appeal of the food wholistically (Experiment 1) and make the nugget cheerleader effect more robust (Experiment 2).In addition, a nugget of a specific shape was rated as the least visually appealing Y. Nishida et al. (Experiments 1 and 2); despite that, it was rated as more appealing when it was presented alongside various shapes (Experiment 2).
Indeed, the lower visual appeal of the ball-shaped nugget was consistently observed despite round objects or foods typically being evaluated favorably.Thus, our results may be an unusual and valuable case.Although this article does not offer a definitive explanation for the results, the shape-taste crossmodal correspondence supplies a plausible explanation.
Fig. 1.Shape of nuggets used for the shape rating task in Experiment 1 Note.The shapes of the nuggets in panels A-D are ball, bell, boot, and bone, respectively.

Fig. 2 .
Fig. 2. Example of rating visual appeal and results of the variety rating task in Experiment Note.Panel A: Participants rated the visual appeal of each plate by placing those that looked tasty on the right and those that looked less tasty on the left.For a larger image, refer to Supplementary Fig. 2. Panel B: Each dot in each condition represents an individual participant's response.The higher the value of visual appeal (yaxis), the tastier the nuggets appeared to the participant.The x-axis indicates the number of nugget shape types on each plate.Error bars indicate SE (N = 21).*p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 3 .
Fig. 3. Comparison of visual appeal against the number of different shapes in the with-and without-ball conditions for Experiment 1 Note.The left side refers to the with-ball condition, and the right side refers to the without-ball condition.The x-axis indicates the number of shape types.Each dot represents an individual participant's averaged response.The higher the value of visual appeal (y-axis), the tastier the nuggets appeared to the participant.Error bars indicate SE (N = 21).There is no four-shape for the without-ball condition.

Fig. 4 .
Fig. 4. Stimuli and trial series of Experiment 2 Note.Panel A: Stimuli for the group conditions of Experiment 2. From left to right: Ball, bell, boot, bone, and all image types (i.e., element nugget shape).Panel B: Series of trials for the group condition.An arrow was presented 1000 ms after the stimulus presentation.Participants rated the nugget at the end of the arrow.Panel C: Series of trials in the single condition.The size of the image on the display was adjusted to be the same as that of the nugget in the group condition.The arrow was not presented in this condition.

Fig. 5 .
Fig. 5. Means of visual appeal in conditions of presentation and element nugget shape types for Experiment 2 Note.Each dot in each condition represents an individual participant's averaged response.The higher the value of the visual appeal score (y-axis), the tastier the nuggets appeared to the participant.The x-axis indicates the element nugget shape type.Error bars indicate SE (N = 22).

Fig. 6 .
Fig. 6.Means of visual appeal in conditions of presentation and target nugget shape types for Experiment 2 Note.Each dot in each condition represents an individual participant's averaged response.The higher the value for visual appeal score (y-axis), the tastier the nuggets appeared to the participant.The x-axis indicates the target nugget shape type.Error bars indicate SE (N = 22).

Table 2
Rated visual appeal, roundness, and aspect ratio for each shape in Experiment 1.