The Delboeuf Illusion affects perceptions of the relative sizes of concentric shapes. This study was designed to extend research on the application of the Delboeuf illusion to food on a plate by testing whether a plate’s rim width and coloring influence perceptual bias to affect perceived food portion size.
Design and methods:
Within-subjects experimental design. Experiment 1 tested the effect of rim width on perceived food portion size. Experiment 2 tested the effect of rim coloring on perceived food portion size. In both experiments, participants observed a series of photographic images of paired, side-by-side plates varying in designs and amounts of food. From each pair, participants were asked to select the plate that contained more food. Multilevel logistic regression examined the effects of rim width and coloring on perceived food portion size.
Experiment 1: participants overestimated the diameter of food portions by 5% and the visual area of food portions by 10% on plates with wider rims compared with plates with very thin rims (P<0.0001). The effect of rim width was greater with larger food portion sizes. Experiment 2: participants overestimated the diameter of food portions by 1.5% and the visual area of food portions by 3% on plates with rim coloring compared with plates with no coloring (P=0.01). The effect of rim coloring was greater with smaller food portion sizes.
The Delboeuf illusion applies to food on a plate. Participants overestimated food portion size on plates with wider and colored rims. These findings may help design plates to influence perceptions of food portion sizes.
Small changes in the food environment can alter perceptions of food size, influence food choices and reduce how much food people serve themselves and consume.1 These environmental changes include, but are not limited to, alterations in portion size,2, 3, 4, 5, 6 container size,7, 8, 9, 10, 11 food variety12, 13 and visual cues about the amount consumed.8 As a result, it may be possible to develop environmental interventions that bypass conscious, cognitive self-control and, instead, ‘passively’ or ‘mindlessly’ produce consistent reductions in energy intake and weight loss over time.
Plate size is one environmental factor that researchers have focused on because it may be an easy factor to manipulate. In the last three decades, the size of the average dinner plate increased more than a third (36%).14 Observational data suggest that the increase in the average size of plates used at meals is concurrent with the rise in obesity rates in the United States.15 Research using mathematical modeling indicates that a small increase in dishware size can lead to a substantial increase in energy intake, particularly if the food is energy-dense.16 Experimental studies show that larger plates make a given portion size look smaller.17 However, smaller plates did not reduce food intake at meals in laboratory studies.18 It is possible that reducing total plate size may be insufficient to reduce intake because substantially smaller plates may make people more aware of smaller portion sizes, which might subsequently result in caloric compensation. However, not all same-sized plates are equal. It may be necessary to pay more attention to plate design characteristics, such as plate color and rim design.
Plate designs that make use of visual illusions, such as the Delboeuf illusion, may independently affect perceived portion size and subsequent food intake. The Delboeuf illusion demonstrates that a circle appears larger when surrounded by a slightly larger circle but smaller when surrounded by a much larger circle19 (Figure 1). The proximity and the color contrast of the concentric circles affect the illusion.20, 21, 22, 23 Research suggests that when the space between concentric circles is small (right image of Figure 1), the circles appear to be one whole object.24 Therefore, people are more likely to emphasize the similarities between the two circles and perceive the inner circle as looking larger than it actually is.25 Conversely, when the space between concentric circles is large (left image of Figure 1), the circles are perceived as two different objects. Therefore, people will emphasize the differences between the two circles,23 leading the inner circle to be perceived as smaller than it actually is.26 In the case of food on a plate, a portion of the food may be perceived as larger on a small plate but smaller on a large plate because of the space between the food and the edge of the plate. Increasing color contrast between the circles has also been shown to enhance the differences between concentric circles.22 Manipulating rim coloring may highlight the edge of the outer circle and color contrast leads people to perceive the size of the inner circle as larger than it actually is22, 27, 28 and perceive a larger food portion size when the color of the plate is different from the color of the food.
Recent research by Van Ittersum and Wansink17 illustrates that the Delboeuf illusion influences serving behavior as the diameter of dishware increases and when there is a color contrast between food and dishware. Big bowls (larger diameter) lead people to serve more food and small bowls (smaller diameter) lead people to serve less food. People also serve themselves more food when food and plates had similar coloring (low contrast). Conversely, people serve themselves less food when food and plates were different colors (high contrast).
The current study extends previous research on how the Delboeuf illusion applies to food on a plate. As the rim of a round plate can be conceptualized as an outer circle surrounding food, we hypothesized that the Delboeuf illusion would apply to plate rims surrounding food on a plate. We describe two experiments in which we manipulated images of plate rim designs and food portions on plates. Our goal was to determine the effect of rim width and color on the perception of portion size. To do this, we quantified perceived food portion size for each rim design used in the study. Experiment 1 examines the effect of rim width on perceived food portion size. Experiment 2 examines the effect of rim coloring on perceived food portion size. Prior work suggests that the Delboeuf illusion is strongest when the outer circle is only slightly larger and/or the color is more greatly contrasted with the inner circle.20, 21, 22, 23 Therefore, in Experiment 1 we hypothesized that wider rims (and thus the closer the inner edge of the rim surrounds the food) would cause food to appear larger on the same-sized plate (Hypothesis 1). In Experiment 2, we hypothesized that colored rims (highlighting the rim to create contrast) would cause food to appear larger than uncolored rims (Hypothesis 2). Results could help inform a superior plate design to influence perceived portion size.
Materials and Methods
We recruited adult participants (aged 18 years or older) via Amazon’s Mechanical Turk (AMT), an online labor market that allows researchers to run experiments in similar formats to computer simulations. AMT has proven to be a reliable tool to conduct behavioral experiments quickly, with valid data.29 Workers in AMT tend to be more representative of the US population as a whole and exhibit behaviors and biases more similar than participants recruited from traditional university subject pools.30 On AMT, we posted a task that required participants to complete an experiment in exchange for $0.50. The task was only visible to participants who were in the United States and were considered to be good workers, defined as having greater than 95% of their past AMT surveys approved as acceptable for data use by other surveyors.
Both experiments used a within-subject experimental design. In each experiment, participants observed multiple side-by-side photographic images of two plates of identical size (total diameter 600 pixels). In both experiments, we included test stimuli with seven food portion sizes (images were sized to 170, 175, 180, 185, 190, 195 and 200 pixels in diameter) on plates with varying rim characteristics in a random order to quantify perceived food size for each rim design used in the study. Every rim characteristic and food size combination were then compared with a standard stimulus (that is, a referent plate), defined as an image with a 185-pixel-food portion size on a plate and a rim width-to-plate radius ratio of 1/3 with no rim coloring (Figure 2). Varying food portion sizes in addition to plate design characteristics is instrumental to quantifying the effects of rim size and coloring on perceived portion using a psychophysical response curve (described below). The test stimuli showing varying food portion sizes and rim characteristics were presented randomly to the left or right of the standard stimulus and the trials were displayed in random order. From each pair, participants were asked to select the plate that contained more food. Food type (macaroni and cheese or fruit salad) was the same in both images within each trial but varied from trial-to-trial and subject-to-subject.
Experiment 1 consisted of 42 trials (seven depicted food portion sizes × six rim widths; no rim, and 1/8, 1/4, 1/3, 3/8 and 1/2 rim width-to-plate radius proportions). Rim widths were chosen by methodically increasing the rim width-to-plate radius proportion of a plate with no rim until the rim width-to-plate radius proportion was half the size of the plate and no longer judged realistic in a real world setting.
Experiment 2 consisted of 28 trials (seven depicted food portion sizes × four rim colors; plain uncolored rim, solid blue colored rim, colored blue line marking the inner border of rim and colored blue lines around inner and outer borders of rim). All plates (standard and test stimuli) had a standard rim of a 1/3 rim width-to-plate radius proportion. As research shows that increasing color contrast between the circles enhances the differences between concentric circles,22 leading people to perceive the size of the inner circle as smaller than it actually is,22, 27, 28 we chose a dark rim color to create contrast against the white plate, highlighting the inner rim surrounding the food image. We also wanted the plate to look as realistic as possible and chose blue because it is one of the most popular plate colors in dishware catalogs. Rim coloring designs (highlighting the rim width as single line on the inner border; highlighting the rim around the inner and outer borders of the rim; and coloring the entire rim) were chosen to provide visual cues and highlight the multiple concentric circles on a plate with a rim (Refer to images of plates in Figure 3).
Psychophysical model of food portion size
To examine how adults estimate food portion size based on plate rim design, we used a psychophysical research approach31 that is frequently used to describe the relationship between subjective perceptions of objective differences. We created psychophysical response curves based on a participant’s ability to correctly identify a larger food portion size when comparing images of plates with different food portion sizes and/or rim characteristics with a standard food portion size and design. Each point on the curve represents the probability of choosing a depicted food size compared with the standard stimulus for each rim design. The point of equivalence, or where each curve crosses 50% probability, marks the point where participants cannot distinguish the difference in food portion size between the test and the standard stimulus, even when they truly exist, where there is an equal 50% chance that participants pick the standard stimulus versus the test stimulus. Comparing points of equivalence allows one to estimate the magnitudes of over- or underestimating in perceived food portion sizes across the different rim characteristics by examining the horizontal shift of the response curve.
Multilevel logistic regression examined the main effect of rim characteristic on the probability of choosing a depicted food size compared with the standard stimulus across rim characteristics, adjusting for the fixed effects of depicted test stimulus food size, food type, the rim characteristics by food size interaction and a random effect for within person variation. We include the rim characteristics × food size interaction term in the primary analysis model to obtain an unbiased estimate of the main effect of rim characteristics, as recommended in the statistical literature.32, 33, 34, 35 The interaction between rim characteristic and depicted food portion size tested whether the probability of picking the test stimulus (that is, rim characteristic) differed by the depicted food size. When interaction terms were significant, post-hoc analyses explored how food size differed by plate design. In all, the analysis models, food portion sizes and rim characteristics were centered on their standard stimulus values.36 Descriptive statistics were used to calculate the magnitude of the relative difference in choosing the test stimulus using odds ratios with 95% confidence intervals, and the points of equivalence for each rim characteristic across food sizes, with the reference point being the standard stimulus. The Wald χ2 test statistic tested the null hypothesis that there is no significant difference between rim characteristics. All tests were two-sided and were conducted at the 0.05 level of significance. Data were analyzed with version 9.2 of the SAS statistical software package (SAS, Cary, NC, USA).
Experiment 1: the effect of rim width on perceived food portion size
A total of 338 participants (55% male; 74% Caucasian, 10% Asian, 6% Latino, 2% Black and 8% other; mean±s.d., age 29.8±11.0 years) completed 42 trials each.
Figure 3 illustrates the probability of choosing a depicted food size compared with the standard stimulus for each rim width-to-plate radius proportion. Table 1 presents a summary of the multilevel logistic model examining the effect of rim width on perceived food portion size. Depicted food portion size was significant in the model (P<0.0001), confirming that participants are able to discern larger food sizes as larger. Rim width (P<0.0001) was also significant in the model, with participants overestimating food portion sizes on plates with greater rim width-to-plate radius proportions (that is, wider rims), confirming Hypothesis 1. There was a significant rim width by depicted food portion size interaction (P<0.0001), indicating that the effect of rim width on perceived portion size is significantly greater with larger food portion sizes. Food type was not significant in the model (P=0.70).
The odds ratios indicate that when comparing the same food size, people were 1.5 times as likely to pick the widest rim plate as having more food compared with the standard plate. Conversely, people were 0.6 times as likely to pick the plate with no rim as having more food. The point of equivalence values estimate the differences in perceived food portion sizes between plates with different rim widths and indicated that there is a 5.2% difference in perceived diameter of the food image when food is viewed on a plate with a very wide rim (1/2 rim width-to-plate radius proportion) compared with a plate with a very narrow rim (1/8 rim width-to-plate radius proportion).
Experiment 2: the effect of rim coloring on perceived food portion size
A total of 251 participants (51% female; 75% Caucasian, 8% Asian, 6% Latino, 6% Black and 5% other; mean±s.d., age 30.0±10.3 years) completed 28 trials each.
Figure 4 illustrates the probability of choosing a depicted food size compared with the standard stimulus for each rim coloring. Table 2 presents a summary of the multilevel logistic model examining the effect of rim coloring on perceived food portion size. Depicted food portion size was significant in the model (P<0.0001), again confirming that participants are able to discern larger food sizes as larger. Rim coloring (P=0.01) was also significant in the model with participants overestimating food portion sizes on plates with any type of rim coloring, confirming Hypothesis 2. There was a significant rim coloring by depicted food portion size interaction (P=0.0003), indicating that the effect of rim coloring on perceived portion size is greater with smaller food portion sizes. Food type was not significant in the model (P=0.47).
The odds ratios indicate that with small food sizes, people were 2.3 times as likely to pick the rims with solid blue coloring as having more food compared with a plate with no rim coloring. However, with larger food sizes, people were only 0.6 times as likely to pick rims with solid blue coloring as having more food compared with a plate with no rim coloring. The point of equivalence indicated that there is a 1.5–1.6% overestimation in the diameter of food portion size when food is viewed on a plate with rim coloring compared with a plate without rim coloring.
We examined whether a plate’s rim design, namely rim width and rim coloring affect, perceived food portion size. Confirming that the Delboeuf illusion applies to plate rims surrounding food on a plate, these experiments found that wider and colored rims exaggerate perceptions of the amount of food on a plate, making images of food look larger. Food type had no effect on the illusion, suggesting that the illusion persists regardless of the type of food depicted on the plate. The effect of rim width is greater with larger food portion sizes consistent with our expectations from the Delboeuf illusion. However, the effect of rim coloring is greater with smaller food portion sizes. The color contrast illusion was inverted when food sizes were larger, resulting in rim coloring making food appear smaller on a plate.
The psychophysical response curve illustrated not only that wider rims and rim coloring exaggerate perceived food portion size but also illustrate how much more food is perceived on the plate. In Experiment 1, the point of equivalence illustrated that food on plates with wider rims looks ∼5% larger in diameter than food on plates with a very thin rim. This difference in diameter is equivalent to a two-dimensional area of 10%, which would represent an even greater exaggeration in terms of three-dimensional food volume. In Experiment 2, food on plates with colored rims looks ∼1.1% larger in diameter, equivalent to an exaggeration of 3% larger in area compared with food on plates with no rim coloring. Putting this into context, if these differences in visual perception translate into adults serving themselves less and eating less, as they perceive more food on plates because of rim design, the differences between eating off a wider rim versus a smaller rim plate could result in clinically important reductions in energy intake.37, 38
Besides the study by Van Ittersum and Wansink17 that tested the Delboeuf illusion with respect to plate size and plate coloring and serving behavior, and a study that tested the illusion on irregular objects, almost all past research on the Delboeuf illusion has been applied to circles and squares. This study extends this prior research on plate size by testing the application of the Delboeuf illusion to images of food and plate design characteristics. Jaeger and Long20 had participants who judged the size of a letter, either an ‘A’ or an ‘S’, which was surrounded by a circle. Similar to our experiments, Jaeger found that an outer circle resulted in people perceiving the letter size as larger. Larger surrounding circles (more space between letter and outer circle, similar to having no rims) diminished the illusion, making the letters look smaller. This study also found that decreasing the lightness contrast (similar to having no rim coloring) of the surrounding circle relative to the central letter diminished the illusion.
Our experiments examined the Delboeuf illusion using two-dimensional images of food and plates, rather than with real food and plates. The use of computers to present visual images of food limits the direct translation of our results to a real world setting. However, the effect of the Delboeuf illusion from a plate’s rim width and coloring may be even greater when taking volume into account. The systematic methods of measuring food portion size and rim widths via visual images on a computer screen also provide a precise demonstration of the illusion in a way that reduces measurement errors. As pictures were shrunk or enlarged to change the size of the food portion, it is possible that participants noticed the subtle differences in the size of the macaroni or fruit, potentially introducing a product-confound. However, we think this is unlikely because we purposely picked amorphous food images in which small changes in the size of components of the images would be difficult to detect. No participants brought up these visual differences during extensive formative and pilot testing. Nevertheless, future studies are needed with real plates and food to validate these findings, to examine the effect of plate design on perceived food amounts and actual energy intake.18
Previous research demonstrates that visual cues of portion size affect serving behavior and food intake. Wansink et al.8 found that people eat more from self-refilling bowls than normal soup bowls. Adults also serve and eat more when given larger bowls and larger serving spoons.39 Research suggests that people cannot overcome these biases through awareness, knowledge or education.7 This suggests that even when people know about the bias created by packaging and portion sizes, it does not protect one from its effect. Therefore, if the findings from our experiment translate to real food on real plates, then plate design may represent a promising weight control strategy.
The strength of these experiments is the large sample size in each experiment, which enhances the generalizability of these findings. As visual perception and perceptual bias are considered relatively hardwired and non-conscious,9 past studies that investigated the Delboeuf illusion were conducted in smaller samples. The present findings also serve as an important first step to conduct this experiment in a real world setting, letting us define important parameters to inform the optimal plate design to influence perceived food portion size and, possibly, subsequent intake.
In conclusion, these experiments tested the Delboeuf illusion with food on plates with different rim characteristics. The results demonstrate that wider and colored plate rims exaggerate the perceived amount of food on a plate, making food look larger, compared with smaller rims, no rims or uncolored rims. The results of these experiments suggest that small changes in plate design have the potential to produce relatively larger effects on perceived food amounts and thus intake. If demonstrated to produce similar effects on actual food portion estimation and intake, plate design utilizing the Delboeuf illusion has the potential to be a simple and important weight loss and maintenance strategy.
Wansink B . Environmental factors that increase the food intake and consumption volume of unknowing consumers. Annu Rev Nutr 2004; 24: 455–479.
Fisher JO, Kral TV . Super-size me: portion size effects on young children's eating. Physiol Behav 2008; 94: 39–47.
Rolls BJ . The supersizing of America: portion size and the obesity epidemic. Nutr Today 2003; 38: 42–53.
Rolls BJ, Morris EL, Roe LS . Portion size of food affects energy intake in normal-weight and overweight men and women. Am J Clin Nutr 2002; 76: 1207–1213.
Ello-Martin JA, Ledikwe JH, Rolls BJ . The influence of food portion size and energy density on energy intake: implications for weight management. Am J Clin Nutr 2005; 82: 236S–241S.
Diliberti N, Bordi PL, Conklin MT, Roe LS, Rolls BJ . Increased portion size leads to increased energy intake in a restaurant meal. Obes Res 2004; 12: 562–568.
Wansink B, Cheney MM . Super bowls: serving bowl size and food consumption. JAMA 2005; 293: 1727–1728.
Wansink B, Painter JE, North J . Bottomless bowls: why visual cues of portion size may influence intake. Obes Res 2005; 13: 93–100.
Folkes V, Matta S . The effect of package shape on consumers’ judgments of product volume: attention as a mental contaminant. J Cons Res 2004; 31: 390–401.
Wansink B . Can package size accelerate usage volume? J Marketing 1996; 60: 1–13.
Wansink B, Kim J . Bad popcorn in big buckets: portion size can influence intake as much as taste. J Nutr Educ Behav 2005; 37: 242–245.
McCrory MA, Fuss PJ, McCallum JE, Yao M, Vinken AG, Hays NP et al. Dietary variety within food groups: association with energy intake and body fatness in men and women. Am J Clin Nutr 1999; 69: 440–447.
Rolls BJ, Rowe EA, Rolls ET, Kingston B, Megson A, Gunary R . Variety in a meal enhances food intake in man. Physiol Behav 1981; 26: 215–221.
Wansink B . Mindless Eating: Why We Eat More Than We Think. Bantam Dell: New York, NY, USA, 2006.
Young LR, Nestle M . The contribution of expanding portion sizes to the US obesity epidemic. Am J Public Health 2002; 92: 246–249.
Pratt IS, Croager EJ, Rosenberg M . The mathematical relationship between dishware size and portion size. Appetite 2012; 58: 299–302.
Van Ittersum K, Wansink B . Plate size and color suggestibility: the Delboeuf Illusion’s bias on serving and eating behavior. J Consumer Res 2012; 39: 215–228.
Rolls BJ, Roe LS, Halverson KH, Meengs JS . Using a smaller plate did not reduce energy intake at meals. Appetite 2007; 49: 652–660.
Delboeuf FJ . “Note Sur Certaines Illusions d’Optique: Essai d’une Théorie Psychophysique de la Manière don’t l’Oeil Apprécie les Distances et les Angles [Note on Certain Optical illusions: Essay on a Phychophysical Theory Concerning the Way in Which the Eye Evaluates Distances and Angles]”. Bulletins de l’Académie Royale des Sciences, Lettres et Beaux-arts de Belgique 1865; 19: 195–216.
Jaeger T, Long S . Effects of contour proximity and lightness on Delboeuf illusions created by circumscribed letters. Percept Mot Skills 2007; 105: 253–260.
Surkys T, Bertulis A, Bulatov A . Delboeuf illusion study. Medicina (Kaunas) 2006; 42: 673–681.
Weintraub DJ, Cooper LA . Coming of age with the Delboeuf illusion: brightness contrast, cognition, and perceptual development. Develop Psychol 1972; 6: 187–197.
Weintraub DJ, Wilson BA, Greene RD, Palmquist MJ . Delboeuf illusion: displacement versus diameter, arc deletions, and brightness contrast. J Exp Psychol 1969; 80: 505–511.
Morinaga S . Conditions of the size-assimilation and size-contrast. In Hirose K (ed.) Masuda Hakushi Shaon Saikin Shinrigaku Ronbunsyuu. Tokoy: Iwanami,, 1935. pp 28–48.
Girgus JS, Coren S . Assimilation and contrast illusions: differences in plasticity. Percep Psychophys 1982; 32: 555–561.
Pollack RH . The effects of fixation upon the apparent magnitude of bounded horizontal extent. Am J Psychol 1964; 77: 177–192.
Oyama T . The effect of hue and brightness on the size-illusion of concentric circles. Am J Psychol 1962; 75: 45–55.
Jaeger T, Lorden R . Delboeuf illusions: contour or size detector interactions? Percept Mot Skills 1980; 50: 376–378.
Rand DG . The promise of mechanical turk: how online labor markets can help theorists run behavioral experiments. J Theor Biol 2011; 299: 172–179.
Paolacci G, Chandler J, Ipeirotis PG . Running experiments on Amazon Mechanical Turk. Judg Decis Making 2010; 5: 411–419.
Stevens SS . Psychophysics: Introduction to its Perceptual, Neural, and Social Prospects. Transaction Publishers: New Brunswick, NJ, USA, 1975.
Cronbach LJ, Snow RE . Aptitudes and Instructional Methods: a Handbook for Research on Interactions. Ardent Media: New York, NY, USA, 1981.
Finney JW, Mitchell RE, Cronkite RC, Moos RH . Methodological issues in estimating main and interactive effects: examples from coping/social support and stress field. J Health Soc Behav 1984; 25: 85–98.
Rogosa D . Comparing nonparallel regression lines. Psychol Bull 1980; 88: 307.
Overall JE, Lee DM, Hornick CW . Comparison of two strategies for analysis of variance in nonorthogonal designs. Psychol Bull 1981; 90: 367.
Kraemer HC, Blasey CM . Centring in regression analyses: a strategy to prevent errors in statistical inference. Int J Methods Psychiatr Res 2004; 13: 141–151.
Hill JO, Wyatt HR, Reed GW, Peters JC . Obesity and the environment: where do we go from here? Science 2003; 299: 853–855.
Hall KD, Sacks G, Chandramohan D, Chow CC, Wang YC, Gortmaker SL et al. Quantification of the effect of energy imbalance on bodyweight. Lancet 2011; 378: 826–837.
Wansink B, van Ittersum K, Painter JE . Ice cream illusions bowls, spoons, and self-served portion sizes. Am J Prev Med 2006; 31: 240–243.
This research was supported, in part, by Award Numbers R01HL096015 and U01HL103629 from the National Heart, Lung, and Blood Institute and the Children’s Health Research Institute at the Stanford University. Dr McClain was supported by Public Health Service Training Grant 5 T32 HL 007034 from the National Heart, Lung, and Blood Institute.
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, And Blood Institute, the National Institutes of Health, or Stanford University.
The authors declare no conflict of interest.
Study design (AM, WV, DM, SM and TR), data collection (AM and WV), data interpretation (AM, WV, DM, MD, SM and TR), literature search (AM and TR), generation of figures (AM, WV, MD and SM), writing of the manuscript (AM, WV, DM, MD, SM and TR).
About this article
Cite this article
McClain, A., van den Bos, W., Matheson, D. et al. Visual illusions and plate design: the effects of plate rim widths and rim coloring on perceived food portion size. Int J Obes 38, 657–662 (2014). https://doi.org/10.1038/ijo.2013.169
- portion size
- visual illusions
- mindless eating
- perceptual bias