An exploratory study on portion size estimation errors: psychological correlates of portion size estimation errors in a non-clinical sample with normal-weight Authors: Mayron Piccolo, Tanya Tandon, Dany Laure Wadji, Stephanie Haymoz, Christophe Mueller-Pfeiffer, Gabriella Milos, Chantal Martin-Soelch Published: 2026 ABSTRACT PurposeAccurate perception of food portion sizes is implicated in healthy eating, yet little is known about psychological correlates of portion estimation in non-clinical, normal-weight adults. This exploratory study examined whether consumption intent (self vs other) and markers of psychological processes associated with eating (dietary restraint, shape/weight overvaluation, body dissatisfaction) are associated with portion size estimation errors, and whether these errors relate to body mass index (BMI).MethodsSeventy-one adults (BMI 18.5–24.9 kg/m2) without current or past psychiatric disorders were included. Eating psychopathology was assessed using the brief 7-item, 3-factor version of the Eating Disorder Examination Questionnaire (Dietary Restraint, Shape/Weight Overvaluation, Body Dissatisfaction). Portion estimation was measured with the Portion Estimation Task, with images rated on a visual analog scale (0–100) under two conditions: general intent (for others) and intent-to-eat (for self). A proportional estimation error index was calculated across all non-full portion sizes relative to each participant’s full-portion reference. A linear mixed model tested the effect of intent and Pearson correlations examined the association with psychological variables and BMI.ResultsConsumption intent did not influence portion size estimation (p = 0.989), so conditions were combined. Estimation Error Indices were positively correlated with Dietary Restraint (r(69) = 0.27, p = 0.021), but not with Shape/Weight Overvaluation or Body Dissatisfaction. BMI correlated with Shape/Weight Overvaluation and Body Dissatisfaction, but not with Dietary Restraint or the Estimation Error Index.ConclusionIn normal-weight adults without eating disorders, greater dietary restraint is associated with larger biases in portion size estimation, whereas BMI and body-related evaluative concerns primarily track weight status rather than perceptual errors in portion size estimations. Findings are exploratory and warrant replication in larger, more diverse samples.Level of evidence: III. Evidence obtained from well-designed cohort or case–control analytic studies. Introduction Accurate estimation of food portion sizes may be an important determinant of healthy eating and weight management among adults. For instance, research has consistently shown that exposure to larger portions increases energy intake [1, 2], which over time may contribute to weight gain [3, 4]. On the other hand, misestimations of food portion sizes are associated with disordered eating, in particular anorexia nervosa [5, 6] and bulimia nervosa [7], as well as in obesity [8]. However, much less is known about how psychological factors associated with eating relate to portion size estimation in non-clinical, normal-weight adults. With the prevalence of obesity among adults in some countries ranging between 35.5 and 46.4% [9], and among children and adolescents worldwide at 14.8% [10], and the prevalence of eating disorders significantly increasing over the past years [11], it is essential to identify factors that may contribute to general weight management, particularly in non-clinical samples, as it can inform interventions aimed at improving dietary habits and preventing both obesity and eating disorders, and help identify subtle psychological processes that may precede clinically significant disturbances. Therefore, the goal of the present study was to investigate factors associated with errors in portion size estimation (Fig. 1).Fig. 1... ... Factors influencing portion size estimations Multiple factors influence portion size estimation, and food characteristics seem to directly influence how individuals estimate meal portion sizes. Almiron-Roig, Solis-Trapala, Dodd and Jebb [12] reported that, in general, participants tend to underestimate portion sizes when presented with real-life food, particularly when the stimulus is composed of one single item (such as an orange juice carton) compared to those with more than one (e.g., cheese and crackers). Additionally, the number of calories (energy density) also differentially impacted food portion estimation, with high-calorie foods leading to overestimation (i.e., considering a portion larger than it is), and medium-calorie foods leading to underestimation of portion sizes. In summary, these findings suggest that food characteristics matter when estimating food portion sizes. Individual characteristics also influence portion size estimation. For instance, Frobisher and Maxwell [13] investigated different methods of portion estimation and identified that children tend to misestimate sizes more than adults, although this difference decreases when images relevant to that age group are used. Almiron-Roig and collaborators [12] also highlighted that females tend to be better at estimating portion sizes than males, although Martin, Han, Coulon, Allen, Champagne and Anton [14], also using real-life meals from participants, found that males outperformed females. These apparent contradictory findings may highlight the need for more controlled designs, including behavioral tasks that use multiple types of stimuli. Notably, despite the relationship between body mass index (BMI) and food portion, Martin et al. [14] reported no significant relationship between body weight and food portion estimation. Eating disorder status also significantly impacts portion size estimations, particularly by leading to errors. Women with anorexia nervosa, for instance, tend to overestimate portion sizes, particularly for a full-size portion and portions smaller than that (e.g., 0.125, 0.25, and 0.5 of a full portion) [5]. This difference becomes more prominent when consumption intent (i.e., who that portion is for) is explored. In other words, overestimation in anorexia nervosa becomes more pronounced when individuals are asked to estimate portions they would eat themselves (intent-to-eat) vs estimating portions that others (someone else) would consume (general intent). In contrast, a small sample of healthy individuals (n = 27) did not show significant differences in portion size estimations when comparing self vs other consumption, suggesting that consumption intent does not lead to bias among this population. Interestingly, across groups, Milos, Kuenzli, Martin-Soelch, Schumacher, Moergeli and Mueller-Pfeiffer [5] identified that drive for thinness was significantly associated with portion size estimations. Pasi, Piccolo, Kaufmann, Martin-Soelch, Muller-Pfeiffer and Milos [7] found similar differences comparing portion estimation in individuals with bulimia nervosa and individuals with normal weight, with the former overestimating portion sizes, similar to individuals with anorexia nervosa. Interestingly, individuals with bulimia nervosa showed smaller error when compared to those with anorexia nervosa. As in Milos et al. [5], eating pathology (e.g., drive for thinness) was a predictor of portion estimation, while BMI was not a significant predictor. Although portion size estimations have been linked to overweight, obesity, and eating disorders in prior research, the present study does not aim to examine obesity directly. Instead, we focused on a rigorously screened sample of individuals with normal weight to investigate whether psychological processes associated with eating psychopathology influence portion estimation errors in the absence of clinically significant weight or psychiatric disorders. More specifically, our goal was to investigate whether consumption intent led to differences in portion estimation. Additionally, we aimed to explore the associations between these psychological processes associated with eating, BMI, and errors in portion size estimation. Based on prior literature, we expected that psychological processes associated with eating (such as dietary restraint, shape and weight overvaluation, and body dissatisfaction), but not consumption intent nor BMI, would be associated with errors in portion estimation. However, because no prior study has examined these relationships in individuals without eating disorders, we considered our analyses exploratory and hypothesis-generating. Examining portion size estimation in adults with normal weight and no significant mental disorders allows for the investigation of early psychological correlates without the confounding influence of clinically significant weight or eating disorders. Methods Participants Participants (n = 71) were recruited from the general population through public announcements at the University of Fribourg in Switzerland. Inclusion criteria for the present study were a body mass index (BMI) between 18.5-24.9 kg/m2, being older than 18 years of age, and not presenting clinically relevant psychiatric symptoms (including any present or history of mental disorders), measured by a standardized clinical interview developed for the present project. The interview assessed participants’ mental health history, medication, substance use, and symptoms of major psychiatric disorders (e.g., mood disorders, anxiety and panic, phobias, and eating disorders. See Table 1 for a detailed description of the present sample and Fig. 1 for the participant inclusion flowchart.Table 1Sample descriptionVariableParticipantsCarrard et al. [17]N (female)71 (46)161 (161)Age in years, mean (SD)22.6 (6.26)28.1 (8.1)BMI, kg/m2 (SD)21.3 (1.73)21.0 (2.3)Clinical questionnaire, mean (SD) Dietary restraint0.967 (1.24)0.9 (1.3) Shape/Weight overvaluation1.62 (1.64)0.9 (1.3) Body dissatisfaction2.18 (1.79)1.7 (1.6)BMI body mass index, SD standard deviation Table 1Sample descriptionVariableParticipantsCarrard et al. [17]N (female)71 (46)161 (161)Age in years, mean (SD)22.6 (6.26)28.1 (8.1)BMI, kg/m2 (SD)21.3 (1.73)21.0 (2.3)Clinical questionnaire, mean (SD) Dietary restraint0.967 (1.24)0.9 (1.3) Shape/Weight overvaluation1.62 (1.64)0.9 (1.3) Body dissatisfaction2.18 (1.79)1.7 (1.6)BMI body mass index, SD standard deviation Sample description BMI body mass index, SD standard deviation Instruments Eating Disorder Examination Questionnaire (EDE-Q)—Brief 7-item, 3-factor version. Markers of eating psychopathology were assessed with selected items from the Eating Disorder Examination Questionnaire [15], using the brief 7-item, 3-factor model proposed by Grilo, Crosby, Peterson, Masheb, White, Crow, Wonderlich and Mitchell [16]. This model yields three subscales: Dietary Restraint (i.e., restraint over eating, food avoidance, dietary rules), Shape/Weight Overvaluation (i.e., importance of weight, and importance of shape), and Body Dissatisfaction (i.e., dissatisfaction with weight, and dissatisfaction with shape). Items are rated on the standard 0–6 EDE-Q response scale (0 = absence of the feature, 6 = feature present every day) asking participants to reflect on the past 28 days. Each score is computed as the mean of its constituent items, with higher scores indicating greater symptom severity. The choice of this briefer, construct-focused scoring approach was informed by Carrard, Rebetez, Mobbs and Van der Linden [17], who examined the factor structure of the EDE-Q in French-speaking Swiss women with and without eating disorders and reported limitations of the original four-subscale structure and superiority of the three-factor version in the non-clinical Swiss context. In the present sample, internal consistency for all scales was good (Dietary Restraint α = 0.858; Shape/Weight Overvaluation α = 0.816, and Body Dissatisfaction α = 0.843). Portion Estimation Task. Portion estimation was measured using a computerized Portion Estimation Task [5, 7]. Participants were presented with meal pictures of different portion sizes and were then asked to indicate the size of each portion using a visual analog scale (VAS; 0 = small and 100 = large). Meal pictures consisted of breakfast and two main courses adapted to the Swiss context and were presented in six different serving sizes (0.125, 0.25, 0.5, 1, 1.5, and 1.75 portions) on a standard plate. Breakfast and lunch were composed of several items, whereas dinner consisted of one single unit with several items composing it (lasagna). One single picture was presented per trial. Images were presented under two consumption intent conditions: general intent (estimate portion for others) and intent-to-eat (estimate portion for self). For each condition, a question was displayed at the top of the screen, asking individuals to rate the portion size if they or someone else was supposed to eat it. Each picture was presented twice: once in the general intent condition and another time in the intent-to-eat condition, with a total of 18 trials per condition. The order of conditions and pictures was randomized. Meal portion sizes were established as directed by the guidelines of the Swiss Society for Nutrition, making the pictures relevant to the current sample. To control for difficulties in perceiving portion sizes, participants were asked to rank pictures from smallest to largest at the end of the trial and were excluded from the study if they made any mistakes (i.e., no mistakes were permitted). No participants were excluded from the study on that basis. Procedure After signing informed consent, all participants were interviewed by a trained graduate student in clinical and health psychology using a brief structured clinical screening interview developed for the present study to ensure the absence of current or recent psychiatric disorders that could affect eating behavior. The interview assessed presence and history of mental disorders (i.e., mood disorders, anxiety and related disorders, substance use disorders, and eating disorders) and current treatment (e.g., lifetime diagnosis of psychiatric disorders, current, or recent (within the past 12 months) symptoms, current psychotropic medication use, and ongoing psychotherapy). Participants with a current psychiatric diagnosis and current treatment linked to a diagnosed mental disorder (psychotropic or psychotherapy) were excluded and, due to our data retention policy, had their information deleted. Following inclusion, all participants completed demographic information, the Eating Disorder Examination Questionnaire [15], and the Portion Estimation Task for food portion estimation assessment [5, 7]. The study was conducted in accordance with Good Clinical Practice and the Declaration of Helsinki for research involving human participants. The study protocol was approved by the Commission Cantonale d’Éthique de la Recherche sur l’Être Humain (CER-VD 2017-01700) and the Institutional Review Board of the Department of Psychology at the University of Fribourg (334-R). Data analysis Consumption intent and portion estimation. To investigate whether consumption intent leads to differences in portion estimation, a linear mixed model was fitted using portion estimation (from the VAS) as the dependent variable and condition (general intent; intent-to-eat) as fixed factor. Subject was included as random factor in the model. Associations between estimation error, psychological processes associated with eating, and BMI. Pearson correlations were conducted to assess the linear relationship between markers of eating psychopathology (Dietary Restraint, Shape/Weight Overvaluation, and Body Dissatisfaction) and errors in portion size estimations (estimation error index; see below for a description of how this measure was calculated) following visual inspection of normality. Additionally, Pearson correlations were employed to assess the relationship between body mass index (BMI) and the same markers of eating psychopathology. We also ran a Pearson correlation between BMI and the estimation error index to examine their relationship. Because the study was exploratory and the sample size modest relative to the number of potential predictors, we relied primarily on bivariate analyses to minimize overfitting and unstable regression estimates [18]. Finally, following a reviewer’s recommendation to rule out sex-related differences in portion-size estimation, we conducted an independent-samples Welch t-test comparing the estimation error index between female and male participants. Estimation error index calculation. To explore our hypothesis, one single portion estimation error index was extracted from the Portion Estimation Task. To control for individual differences in portion estimation and avoid bias from absolute portion sizes, error indices were calculated relative to each participant’s own mean estimate for the full portion size (size = 1). First, the mean estimated portion (from the VAS scale) was computed for each participant and each portion size, across all trials. For each participant, the mean estimate for the full portion size served as that participant’s personal reference value (“full portion”), representing 100% of their own scale. We then calculated the expected estimate index for each portion size per participant as follows:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{Expected}}_{i} \, = \,\left( {{\mathrm{Participant}}^{\prime}{\mathrm{s}} \,{\mathrm{full}} \,{\mathrm{portion}} \,{\mathrm{mean}}} \right)\, \times \,{\mathrm{size}}_{i}$$\end{document}Expectedi=Participant′sfullportionmean×sizei The error for each portion size was then calculated as the signed (+ or −) proportional error:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{Proportional}}\,{\mathrm{error}}_{i} = \frac{{{\mathrm{Mean}}\,{\mathrm{estimated}}\,{\mathrm{portion}}_{i} \, - \,{\mathrm{expected}}_{i} }}{{{\mathrm{expected}}_{i} }}$$\end{document}Proportionalerrori=Meanestimatedportioni-expectediexpectedi Error values greater than zero indicate overestimation (perceived size greater than proportionally expected), while values below zero indicate underestimation. To summarize overall estimation bias for each participant, we calculated the mean proportional error across all portion sizes except for the full portion (where error is always zero). This resulted in a single error index per participant, such that positive values denote a tendency to overestimate, and negative values reflect underestimation of portion size relative to their own scaling reference. Alternative approaches, such as absolute error scores, may be influenced by individual differences in VAS scaling tendencies. In other words, it assumes that all participants use the scale in a comparable manner, whereas that may not be the case [19]. By indexing errors relative to each participant’s full-portion reference, we aimed to reduce intersubject scaling biases. Statistical significance was set at p < 0.05 and all statistical analyses were done using RStudio (2025.05.1 + 513). Data and coding are available upon request. Results Consumption intent and portion estimation Our linear mixed model showed no evidence that consumption intent influenced portion size estimation (p = 0.989). On average, estimated portion size in the general intent condition was 39.45 (SE = 0.82) on the VAS, and the difference between the intent-to-eat and general intent conditions was negligible (β = 0.01, SE = 0.94, t = 0.01). As no differences were seen between conditions, this was disregarded for the next analyses as in Pasi et al. [7]. Associations between estimation error, markers of eating psychopathology, and BMI Partially as predicted, Estimation Error Indices were significantly correlated with Dietary Restraint (r(69) = 0.27, p = 0.021). The correlations between Estimation Error and Shape/Weight Overvaluation (r(69) = 0.09, p = 0.475) and Body Dissatisfaction (r(69) = 0.16, p = 0.187) were not significant. We further explored the relationship between BMI and markers of eating psychopathology (EDE). BMI was significantly correlated with Shape/Weight Overvaluation (r(69) = 0.28, p = 0.019) and Body Dissatisfaction (r(69) = 0.34, p = 0.004), but not with Dietary Restraint (r(69) = 0.07, p = 0.555). Table 2 contains a summary of correlations.Table 2Correlation tableBMIr (p-level)Estimation error indexr (p-level)Dietary restraintr (p-level)Shape/weight overvaluationr (p-level)Body dissatisfactionr (p-level)BMI−0.01 (0.957)0.07 (0.555)0.28 (0.019)*0.34 (0.004)*Estimation error index−0.01 (0.957)0.27 (0.021)*0.09 (0.475)0.16 (0.187)BMI body mass index Table 2Correlation tableBMIr (p-level)Estimation error indexr (p-level)Dietary restraintr (p-level)Shape/weight overvaluationr (p-level)Body dissatisfactionr (p-level)BMI−0.01 (0.957)0.07 (0.555)0.28 (0.019)*0.34 (0.004)*Estimation error index−0.01 (0.957)0.27 (0.021)*0.09 (0.475)0.16 (0.187)BMI body mass index Correlation table BMI body mass index Finally, no significant correlation was found between the estimation error index and BMI (r(69) = −0.01, p = 0.957). Similarly, no significant between-sex differences emerged (Mfemale = −0.069, SD = 0.240; Mmale = −0.106, SD = 0.255; t(46.92) = 0.59, p = 0.559). Discussion The present study focused on psychological mechanisms underlying portion estimation in a normal-weight, non-clinical sample as potential early markers of clinically significant weight-related outcomes. Because our sample was restricted to normal-weight individuals, conclusions cannot be extended to populations with overweight or obesity. Portion estimation error indices were significantly correlated with dietary restraint, measured by the 3-factor version [16] of the Eating Disorder Examination Questionnaire (EDE-Q) [15]. Furthermore, BMI was significantly correlated with shape and weight overvaluation and body dissatisfaction, but not with dietary restraint. Consumption intent did not modulate portion estimation errors, suggesting this is mostly the case for individuals with anorexia nervosa [5]. Finally, BMI and portion estimation error were not significantly correlated. Importantly, our sample consisted of normal-weight individuals without an eating disorder diagnosis (M BMI = 21.3, SD BMI = 1.73). Portion estimation error was associated with dietary restraint. Although statistically significant, the association between dietary restraint and portion estimation error was modest in magnitude, suggesting that dietary restraint explains only a limited proportion of variance in portion estimation bias. Individuals with high levels of eating concern and food restraint may experience anxiety when presented with food options, which may explain errors in portion estimation. Recent research by Dicker-Oren, Gelkopf and Greene [20] supports this relationship. Using momentary ecological assessment, the authors found that high food restraint was associated with higher levels of state anxiety but not BMI. As in the present study, this relationship was observed in a non-clinical sample. Coupled with Dicker-Oren and collaborators’ [20], our findings suggest that the interplay between dietary restraint and anxiety may, potentially, underlie biases in portion estimation even in healthy normal-weight individuals. Shape and weight overvaluation and body dissatisfaction, in our sample, were associated with BMI but not as strongly to portion size estimations. Calzo, Sonneville, Haines, Blood, Field and Austin [21], for instance, found that higher BMI was significantly associated with weight concern for both males and females, which is in line with findings of the present study. Interestingly, among adolescents, increasing age was significantly associated with increasing shape and weight concern. In summary, these patterns suggest that while shape and weight concerns may be reliable predictors of BMI, they do not appear to effectively explain errors in portion estimation. Of note, it is conceivable that the study participants were individuals who paid particular attention to the topic of nutrition, which may have influenced the results. Strengths and limitations The present study focuses on a non-clinical, normal-weight adult sample rigorously screened for current and past psychiatric disorders, which allows for the examination of portion estimation biases and psychological processes associated with eating in individuals without psychiatric disorders. The restricted BMI range and psychiatric screening allowed us to isolate psychological correlates of portion estimation without confounding from clinical eating disorders or obesity. Second, we used a standardized computerized task based on a real-life meal tailored to meet Swiss nutritional guidelines and developed an error index that controlled for individual scaling differences. Finally, we employed a validated measure of eating psychopathology using a construct-focused, three-factor model of the EDE-Q, which strengthens the construct validity and interpretability of our findings in non-clinical contexts. Some aspects of this study warrant attention. Firstly, although this study was exploratory, the small sample size limits the generalizability of these findings. Additionally, the findings cannot be generalized to populations with eating or weight disorders, as our sample was restricted to individuals with normal weight and no clinically significant psychiatric symptoms. Secondly, although population norms are missing for the Swiss population, our participants’ means for shape/weight overvaluation and body dissatisfaction were relatively higher compared to another study with the same population ([17]; see Table 1 for a comparison). While this is not surprising given that our sample consisted of younger individuals, it may also reflect broader sociocultural influences such as social media use [22]. Exposure to idealized and often digitally altered body images on platforms such as Instagram has been linked to increased body dissatisfaction and shape/weight concerns even in non-clinical samples [22–24]. Such exposure may therefore have contributed to higher levels of body-related concerns in our sample, as there is approximately one decade separating Carrard and collaborators’ [17] and the present study. Thirdly, BMI was derived from self-reported height and weight, which are known to be affected by modest under-reporting of weight even among normal-weight individuals, possibly attenuating associations involving BMI [25]. It is, however, important to note that this trend is often negligible among European samples, which is the case in the present study [26, 27]. Finally, because the analyses were exploratory and hypothesis-generating, no correction for multiple comparisons was applied; therefore, the results should be interpreted as preliminary and requiring replication in larger confirmatory studies. Future studies should continue to explore the relationship between errors in portion estimation and psychological aspects of eating in larger, more diverse samples, and investigate how these associations may predict psychopathology and metabolic outcomes. Identifying psychological factors that contribute to biases in portion size estimation may help clarify the mechanisms underlying eating behavior. In the present study, increased error in portion estimation was associated with higher dietary restraint, a marker of disordered eating [15, 28]. These findings suggest that portion estimation biases may influence how individuals engage with food and could represent early psychological correlates warranting further longitudinal investigation. What is already known on this subject? Larger food portion sizes lead to increased energy intake and are associated with higher body weight.Individuals with eating disorders, particularly anorexia nervosa and bulimia nervosa, tend to overestimate meal portion sizes, and this bias is more related to eating psychopathology than body mass index (BMI).Prior work on portion size estimation has focused mainly on clinical populations or has emphasized food characteristics (e.g., energy density, number of items) and basic individual differences (e.g., age, sex), with less attention to psychological mechanisms in non-clinical, normal-weight adults. Larger food portion sizes lead to increased energy intake and are associated with higher body weight. Larger food portion sizes lead to increased energy intake and are associated with higher body weight. Individuals with eating disorders, particularly anorexia nervosa and bulimia nervosa, tend to overestimate meal portion sizes, and this bias is more related to eating psychopathology than body mass index (BMI). Individuals with eating disorders, particularly anorexia nervosa and bulimia nervosa, tend to overestimate meal portion sizes, and this bias is more related to eating psychopathology than body mass index (BMI). Prior work on portion size estimation has focused mainly on clinical populations or has emphasized food characteristics (e.g., energy density, number of items) and basic individual differences (e.g., age, sex), with less attention to psychological mechanisms in non-clinical, normal-weight adults. Prior work on portion size estimation has focused mainly on clinical populations or has emphasized food characteristics (e.g., energy density, number of items) and basic individual differences (e.g., age, sex), with less attention to psychological mechanisms in non-clinical, normal-weight adults. What this study adds? In healthy, normal-weight adults without a history of psychiatric disorders, greater dietary restraint is associated with greater errors in portion size estimation, suggesting that even subclinical eating psychopathology may be linked to biased perception of meal portions.BMI in this non-clinical sample is more strongly associated with shape/weight overvaluation and body dissatisfaction than with portion estimation error, indicating that body-related concerns may track weight status more closely than perceptual biases in portion size.Consumption intent (estimating portions for self vs others) does not significantly influence portion size estimation in healthy adults, in contrast with patterns observed in individuals with anorexia nervosa. In healthy, normal-weight adults without a history of psychiatric disorders, greater dietary restraint is associated with greater errors in portion size estimation, suggesting that even subclinical eating psychopathology may be linked to biased perception of meal portions. In healthy, normal-weight adults without a history of psychiatric disorders, greater dietary restraint is associated with greater errors in portion size estimation, suggesting that even subclinical eating psychopathology may be linked to biased perception of meal portions. BMI in this non-clinical sample is more strongly associated with shape/weight overvaluation and body dissatisfaction than with portion estimation error, indicating that body-related concerns may track weight status more closely than perceptual biases in portion size. BMI in this non-clinical sample is more strongly associated with shape/weight overvaluation and body dissatisfaction than with portion estimation error, indicating that body-related concerns may track weight status more closely than perceptual biases in portion size. Consumption intent (estimating portions for self vs others) does not significantly influence portion size estimation in healthy adults, in contrast with patterns observed in individuals with anorexia nervosa. Consumption intent (estimating portions for self vs others) does not significantly influence portion size estimation in healthy adults, in contrast with patterns observed in individuals with anorexia nervosa.