The effect of sad facial expressions on weight judgment

Abstract

Although the body weight evaluation (e.g., normal or overweight) of others relies on perceptual impressions, it also can be influenced by other psychosocial factors. In this study, we explored the effect of task-irrelevant emotional facial expressions on judgments of body weight and the relationship between emotion-induced weight judgment bias and other psychosocial variables including attitudes toward obese persons. Forty-four participants were asked to quickly make binary body weight decisions for 960 randomized sad and neutral faces of varying weight levels presented on a computer screen. The results showed that sad facial expressions systematically decreased the decision threshold of overweight judgments for male faces. This perceptual decision bias by emotional expressions was positively correlated with the belief that being overweight is not under the control of obese persons. Our results provide experimental evidence that task-irrelevant emotional expressions can systematically change the decision threshold for weight judgments, demonstrating that sad expressions can make faces appear more overweight than they would otherwise be judged.

Keywords: body weight; emotions; facial expressions; facial perception; obesity.

FIGURE 1

(A) Exemplar facial stimuli used for the weight judgment task. A total of four identities (two male identities and two female identities) were used in the main experiment. Normal weight (0%) images are shown. (B) Emotional expression and weight of facial stimuli were manipulated by using morphing software. Faces have weight gradients ranging from 0% (normal weight) to 100% (highly overweight) by increments of 20%. Neutral and sad faces are the exact same size and only differ in their emotional expressions.

FIGURE 2

(A) Male face data. (B) Female face data. Average probability of fat responses as a function of weight levels (0–100% overweight) and emotional expressions. Error bars denote the SE of the mean. ^∗p < 0.01; ^∗∗p < 0.005; ^∗∗∗p < 0.001. (C) For the weight judgment data, psychometric curves were fitted by using the Naka-Rushton response function. A leftward-shift of a psychometric curve of Male Sad faces (red line) compared to Male Neutral faces (blue line) was observed. A horizontal dotted line represents the 50% probability of fat decision. (D) Scatter plot of the relationship between BAOP (Belief About Obese Persons) scores and C₅₀ differences between Male Neutral faces–Male Sad faces. Higher BAOP scores indicate a stronger belief that obesity is not under the obese person’s control. Solid line represents a linear fit.