1
Physical Appearance, Attractiveness, and the Mediating Role of Emotions
Albert Mehrabian and Jeffrey S. Blum
The present study was designed to investigate effects of most physical features previously shown to be determinants of physical at tractiveness. Findings from studies conducted primarily with college students in the United States helped delineate the physical features to be explored.
Physical correlates of attractiveness have been categorized as static (or stable) and fluctuating (or changeable) features (Brown, Cash, & Noles, 1986). āStaticā features included relatively enduring physical characteristics such as height or eye color. āFluctuatingā or changeable features referred to characteristics that varied over time, such as hair style or facial pleasantness.
Body shape has been found to be an important stable determinant of attractiveness. Shapeliness of physique related to ratings of attractiveness in men and women. Women preferred men exhibiting V-shaped bodies (Lavrakas, 1975). Also, upper body strength enhanced ratings of male attractiveness (Franzoi & Herzog, 1987). In comparison, men favored women with hourglass figures (Gitter et al., 1983).
Obesity has been found to decrease attractiveness (Clayson & Klassen, 1989). More specifically, Gitter, Lomranz, and Saxe (1982) used drawings of male figures varying in abdomen, shoulders, neck, head, and body shapes. A protruding abdomen was the strongest indicator of unattractiveness. In another study, the negative effect of obesity on attractiveness was more pronounced for female than for male targets (Furham & Radley, 1989).
Height also has been shown to be a determinant of attractiveness. Men found shorter women more attractive and more desirable as dates, whereas women preferred to date taller men, but did not rate them as more attractive than shorter men (Shepperd & Strathman, 1989).
Many stable physical properties of the face have been studied also. Overall facial attractiveness was found to correlate positively with attractiveness of its component parts, with the mouth region having the greatest influence, followed by the eyes, hair, nose, and facial structure (Terry & Davis, 1976).
A considerable number of studies have dealt with so-called ābabyish facial features,ā grouping facial cues on the basis of a priori assumptions rather than on the basis of factor-analytic findings. Zebrowitz and Montepare (1992) defined a babyish face as follows: āfacial features of large eyes, thin and high eyebrows, a large cranium, a small chin, and a curved rather than an angular faceā (p. 1143). Berry and McArthur (1985) found that large, round eyes, a small chin, and high eyebrows of menās faces were judged to be babyish in appearance. Furthermore, a composite score, based on the latter babyish characteristics, correlated positively with attributions of naivete, honesty, warmth, and kindness to stimulus persons. Finally, although correlations of babyish facial features with attractiveness were not reported, subjectively judged babyish facial quality and attractiveness of stimulus persons were positively intercorrelated. In a subsequent review, Berry and McArthur (1986) concluded that adults with more babyish facial features were perceived to have more of the following childlike traits: naivete, honesty, warmth, submissiveness, and physical weakness.
Working along similar lines, Cunningham (1986) used three conceptual groupings of facial features (neonate or babyish, mature, and expressive) to study attractiveness. He found that men judged women with larger eyes, a smaller chin, and a smaller nose as more attractive. āMatureā features of more prominent cheekbones and narrower cheeks, and āexpressiveā features of higher eyebrows, larger pupils and a larger smile, were also judged to be more attractive. In a subsequent study, Cunningham et al. (1995) obtained the following corroborating positive correlates of physical attractiveness: larger eyes, wider-set eyes, larger pupils, higher eyebrows, bigger smile, more prominent cheek bones, narrower cheeks and face; also, for men only, less bushy eyebrows and a smaller chin (Table 8). However, a thorough investigation of babyish facial cues across a wide age span showed no relation between babyish facial features and attractiveness, except in the case of infants. Instead, the evidence showed that babyish facial features connoted childlike traits (Zebrowitz & Montepare, 1992, Tables 3 and 4).
It is important to note differences in classification of babyish facial features by Cunningham et al. (1995) and by Zebrowitz and Montepare (1992). Inconsistencies were due to Cunninghamās conceptual differentiations among neonate (babyish), mature, and expressive facial features (e.g., eyebrow height and chin size were not part of Cunninghamās definition of babyish features, whereas these two cues helped define babyish features for Zebrowitz & Montepare). Thus, lack of agreement regarding the definition of babyish facial features and inconsistencies in findings relating these features to attractiveness were two areas of difficulty when the concept of babyish facial features was applied to study physical attractiveness. The present study, although not focused directly on babyish facial features, included data that would (a)yield a statistical test of assumed groupings of babyish facial features and (b) assess relations of such features with attractiveness.
Differential contributions of face and body to overall judgments of attractiveness have also been investigated. In one study that combined three levels each of facial and bodily attractiveness, faces as well as bodies significantly influenced overall ratings (Alicke, Smith, & Klotz, 1986). In another study, the influence of facial attractiveness (27% of total variance) outweighed that of bodily attractiveness (20% of variance) (Mueser et al., 1984).
There are fewer studies dealing with influences of changeable features on judged attractiveness. Brown, Cash, and Noles (1986) found that overall grooming correlated positively with judged attractiveness of both men and women; tighter clothing was judged more attractive for male targets only; and perceived masculinity (femininity) correlated positively with attractiveness of male (female) targets.
Hair color and length have been found to influence attractiveness. Lighter hair was judged as more attractive. Blondes received the highest, and redheads the lowest attractiveness ratings (Clayson & Klassen, 1989; Clayson & Maughan, 1986). Longer and fuller hair were judged as more attractive in women (Cunningham et al., 1995). Women judged shorter hair as more attractive in men, although the relation was reversed for liberal women (Peterson & Curran, 1976).
Other changeable factors such as glasses, wardrobe, and makeup have been found to affect attractiveness as well. Wearing glasses reduced attractiveness ratings, particularly for female targets (Edwards, 1987). More formal attire (ranging from jeans/shirt to slacks/sweater, suits, and, most formal, uniforms) was judged as more attractive (Hewitt & German, 1987). Makeup increased menās ratings of womenās attractiveness, but it did not influence womenās ratings of womenās attractiveness (Cash et al., 1988). Finally, in line with reinforcement theories of attraction and affiliation (Byrne & Clore, 1970; Clore & Byrne, 1974; Mehrabian & Ksionzky, 1970), studies showed positive relations between positive/negative facial expressions and attractiveness (Gouaux, 1971; Mueser et al., 1984; Veitch 1976).
The present study investigated physical features noted in the preceding review. Hypotheses of the study relating attractiveness to specific physical features were derived from findings reviewed above. Because the reviewed studies typically employed U.S. university students, our hypotheses were proposed as being applicable, primarily, to the young adult segment of the U.S. population.
The hypotheses were as follows: For both men and women, physical attractiveness correlates positively with shapeliness of physique, overall grooming, more formal clothing, more positive and/or less negative facial expressions, less obesity, and flatness versus protrusion of abdomen. For men, physical attractiveness also correlates positively with bodily strength, perceived masculinity, and tighter clothing and, for women, it correlates positively with perceived femininity. Due to inconsistencies in findings bearing on relations of babyish facial features with attractiveness, no hypotheses were offered regarding babyish facial features.
Broader objectives of the present study were of a conceptual nature and involved the Pleasure-Arousal-Dominance (PAD) Emotion Model (e.g., Mehrabian, 1995). Precursors of the model were developed in the 1960s during studies of nonverbal communication. To harness the considerable complexity and diversity of nonverbal cues, referents or meanings (e.g., inferences of emotions and attitudes of another based on the otherās nonverbal acts) rather than symbols or discrete behaviors (e.g., movements, postures, voice quality) of nonverbal communication were used to group and understand the significance of nonverbal cues in communication (e.g., Mehrabian, 1972; 1981).
The focus on referents helped rapid identification of coding rules (i.e., relations among symbols, on one hand, and referents, on the other) in nonverbal communication. Also, major referent dimensions in nonverbal communication were found to correspond to the Evaluation, Activity, and Potency factors of the semantic differential (Osgood, Suci, & Tannenbaum, 1957). This led to formulation of a general three-dimensional framework for describing emotions that included pleasure-displeasure (the positive emotional correlate of Evaluation), arousal-nonarousal (the positive emotional correlate of stimulus Activity), and dominance-submissiveness (the negative emotional correlate of stimulus Potency) (e.g., Mehrabian, 1995). Specific emotions were described as weighted linear combinations of the PAD factors (e.g., excitement or triumph included pleasure, high arousal, and dominance; fascination or respect were composed of pleasure, high arousal, and submissiveness; security or relaxation included pleasure, low arousal, and dominance; anger or hostility included displeasure, high arousal, and dominance; despair and boredom involved displeasure, low arousal, and submissiveness).
The PAD Emotion Model seemed particularly suited to study of the highly diverse features identified in research on physical attractiveness. As with research on nonverbal communication, it was expected that study of the emotional impact of various physical features would provide a parsimonious emotion-based description of the attraction process.
In short, the plan of the present study included (a) identification of statistically-founded and, hopefully, meaningful groupings of physical features, (b) assessment of the differential importance of various groups/factors in determining attractiveness, and (c) assessment of emotional response concomitants of physical attraction to another. To explore the role of emotions in physical attraction, emotional reactions of raters were treated as mediating the relation between the independent variables (physical features) and the dependent variable (judged attractiveness of another).
Method
Participants
These were 117 University of California undergraduates (53 men, 64 women) who participated in the study in partial fulfillment of a course requirement.
Materials
Photographs of targets rated for attractiveness. Student volunteers from University of California, Berkeley, and University of California, Los Angeles, and other individual volunteers from the West Los Angeles area were photographed in a standardized format. Ages of the targets ranged from 18 to 32 years. Photographs were taken at a distance of five feet, with the target standing against a white ...