Face-ism and Self-Attributed Face-ism

Face-ism is a quantitative ratio, wherein the distance from the top of the head to the bottom of the chin is divided by the distance from the top of the head to the lowest visible part of the body in the picture (^{Archer et al., 1983}). It is also known as facial prominence or the face/body ratio.

Face-ism was initially studied in photographs from magazines, advertisements, or political campaigns, where people other than those in the pictures chose them (from the 1970s to the 1990s). However, nowadays, with the explosion and popularization of social networks, people choose their own pictures to represent them to the world. For this reason, we will use the term "self-attributed face-ism" (SAF) in this paper to highlight that now people choose their own pictures on SNS. Therefore, this study is not about traditional face-ism, but rather about on-line SAF as a nonverbal, voluntary and unconscious phenomenon performed by SNS users.

The theoretical framework of face-ism has been a complex and controversial topic that has evolved with the emergence of new theories related to gender in recent decades. Social learning theory (^{Bandura & Walters, 1963}) was the most relevant theory to explain SAF. It suggests that such responses are acquired through observational learning, where imitation plays an important role. The symbolic models of men (mind) and women (body) are integrated by observing real-life models or being exposed to social media models and messages (^{Coltrane & Adams, 1997}). Therefore, this phenomenon occurs unevenly in different cultures (^{Cooley & Smith, 2013}). Men and women's symbolic models are critical factors in shaping behavior or social norms, gender patterns, and sex roles (^{Bussey & Bandura, 2004}). Face-ism was later analyzed by observing the power relations between men and women according to the social role theory of sex differences (^{Carli & Eagly, 2001}; ^{Eagly & Steffen, 1984}). Thus, ^{Dodd, Harcar, Foerch, and Anderson (1989}) and ^{Sparks and Fehlner (1986)} suggested that the social role of the sitter is an essential factor in the regulation of facial prominence in a picture, and they showed that women in typically feminine social positions showed a higher percentage of the body. From the objectification theory perspective (^{Fredrickson & Roberts, 1997}), women in Western societies are sexually objectified, and therefore, women's bodies would have a more relevant role in a picture.

Summing up, authors who have previously studied the complex phenomenon of face-ism have suggested various theoretical frameworks related to social learning and imitation of gender roles expressed through body image. However, there is a lack of a theoretical framework to understand the phenomenon when the person performs it, i.e., SAF. In this study, we propose some factors that might be affecting the continuity of the phenomenon on-line.

Self-Attributed Face-ism and gender

SAF differences are related to sexist attitudes, given that facial images are predominantly associated with attributes such as intelligence, ambition, and fierce competition (^{Schwarz & Kurz, 1989}), ambition and dominance (^{Zuckerman & Kieffer, 1994}), or a generally positive assessment (^{Levesque & Lowe, 1999}). In contrast, images with a higher body proportion are associated with the sexual representation of the body and objectification (^{Heflick & Goldenberg, 2014}).

Gender stereotypes associate men as being independent, goal-oriented, brave, and engaged in professional or leisure activities; by contrast, women are represented as emotional, dependent, and unambitious (^{Bussey & Bandura, 2004}). An example of this is the case of female politicians. The study of ^{Jungblut and Haim (2021)} shows that, despite the party lines, female candidates in the 2019 European election among European Union's 28 member states resemble emotional gender stereotypes, especially since women are more often self-depicted as happy in their SNS. Thus, women and men prioritize different values: men consider work and self-promotion to be the most important values, whereas women give priority to mutual interdependence and, therefore, human bonds. This situation occurs in many cultures where there is a dominant group (men) and a subordinate one (women), and stereotypes shown in social media reinforce and perpetuate these differences (^{Lips, 2001}). In this line,

Hypothesis 1. The pictures will present SAF differences between men and women, showing men's higher facial rate.

Self-Attributed Face-ism, gender and age

In recent years, some studies have also included age as a variable when studying on-line profiles, yielding opposite results. On the one hand, some studies show no difference in on-line identities constructed by boys and girls between 18 and 23 years old (^{Hum, Chamberlin, Hambright, Portwood, Schat & Bevan, 2011}), suggesting that gender differences may be starting to dissolve among the younger generations. On the other hand, ^{Smith and Cooley (2012}) described age as a SAF moderator on Facebook. Based on the idea that older generations will keep stronger gender stereotypes, and then, the facial prominence of men pictures over women's ones, we propose:

Hypothesis 2. Age will influence SAF, increasing for older age.

Another critical issue related to SAF remains unexplored due to a lack of previous studies on it. Although age, sex, and SNS are proposed as moderators, no study has analyzed these variables' interaction (^{Hayes, 2018}). Some features of the face-ism phenomenon, such as intelligence, professional competence, or dominance, could be affected by sex in interaction with age, increasing more quickly in men than in women. Thus,

Hypothesis 3. There will be a statistically significant interaction between age and sex to predict SAF; the rate of SAF growth will be different for men than for women, growing more for men.

Specific Social Network Sites

As ^{Li, Chen, Cheng, Sang and Lee (2019)} point out, people use the internet for various purposes (e.g., informational, relational, recreational, etc.). SNS usage includes not only hedonic (pleasure-oriented) but also utilitarian (rational and goal-oriented) gratifications (^{Xu, Ryan, Prybutok & Wen, 2015}). In this line, SNSs different purposes range mainly from more professional and public ones (i.e., LinkedIn) to more private such as erotic dating ones (i.e., Badoo), including in the middle what we could call relational sites or "connecting with friends" ones (i.e., Facebook). Thus, the user's purpose of communication can dictate the selection of self-images (^{Sanderson, 2008}).

We should not forget the role of sexuality when selecting these self-images; body representation in pictures is also affected by personal and social experiences of sexuality. Previous literature reviews on body representation and sexuality are abundant (^{Kuhn, 1985}), but only some emerging scientific articles analyze this topic on SNS. In this sense, objectification of women and sexual inequality are evident in images on male-oriented pornography sites (^{Shim, Kwon & Cheng, 2015}). Young users perceive the use of filters positively to increase their appeal. They want to see attractive images, despite knowing that they are not accurate representations of self and surroundings (^{Djafarova & Trofimenko, 2019}). Then, according to the social role theory, a network's theme and objectives (professional objectives, erotic dating, or connecting with friends) could influence picture representation.

Hypothesis 4. SAF differences between SNS are expected, increasing professional networks (i.e., LinkedIn) and decreasing on erotic-dating networks (i.e., Badoo).

Figure 1 shows a conceptual representation of the hypotheses.

Figure 1 Conceptual representation of the hypotheses 

This study primarily aims to determine the variables that affect face-ism and make three main new contributions to current research on this issue. First, including age as an independent variable with a wide age range (16 to 73 years old). Second, going beyond previous studies on SAF focused on only one SNS, mainly Facebook, we analyze three SNS. And third, considering the interaction of the variables (or the moderation of variables, as other authors call it) age and sex in the different SNS for forecasting SAF.

Sample

A random sample of 1050 profiles corresponding to three different SNS was analyzed: 350 pictures extracted from a general SNS (Facebook), 350 from a professional-oriented SNS (LinkedIn), and 350 ones from an adult-erotic dating SNS (Badoo). Of these, 525 were women's and 525 were men. The subjects' age range was between 16 and 73 years (M = 36.32, SD = 8.96), located in Spain.

Although an SNS profile can contain many pictures, we only measured the main profile picture in this study. We randomly selected SNS profiles with a public photograph, sex, and age, rejecting profiles without a person's picture in the portrait format or non-human photos. For example, we exclude images of more than one person, non-human portraits (animals, landscapes, etc.), and photographs of actions or body parts.

Variables

The Self-attributed face-ism (SAF) ratio is a quantitative measurement expressed as a percentage of the face in a picture (^{Dodd et al., 1989}), as a statistical result is invariant to the unit of measure. SAF was measured in millimetres using a screen measuring tape. For example, in Figure 2 the length of the face is A (5.4), and its showed body length is B (7.4), so the face-ism ratio of this picture is A/B = 5.4/7.4 = .73. This value was multiplied by 100 to make the statistical results more interpretable, resulting in % SAF = 73 (the ratio of the length of the face to the visible part of the body in the portrait is 73%). A high face-ism score indicates a large representation of the face concerning the rest of the body, and vice versa.

Figure 2 Example of Self-attributed Face-ism ratio calculated on a picture, SAF = A/B = 5.4 / 7.4 = .73 

Sex is a nominal variable characterized numerically as a dummy variable (0: men, 1: women). Because it was impossible to measure gender as such a complex attribute, we consider the sex that each person in the sample indicated in their profile.

Age is a numerical scale variable measured in years.

We transformed the Social Network Site (SNS) into dummy variables. Thus, SNS= (Facebook, LinkedIn), with Badoo as the reference group (^{Hardy, 1993}; ^{Hayes, 2018}; ^{Rosel, Jara & Herrero, 2014}).

Data analysis

The data were statistically analyzed (^{SPSS 27}), obtaining descriptive statistics and graphics. We performed a linear regression analysis using age, sex, their interaction (Sex×age), and SNS as independent variables (IVs), and SAF as the dependent variable (DV). Statistical significance was p < .05. Data can be unloaded in Universitat Jaume I repository: http://repositori.uji.es/xmlui/handle/10234/192402.

Theoretical contributions

The contributions of this research are both theoretical and practical. Regarding theory, it is essential to stress the relation of our empirical results with some critical theories argued in previous studies to explain the face-ism phenomenon. Face-ism was first studied mainly in adverts (^{Belkaoui & Belkaoui, 1976}; ^{Furnham & Bitar, 1993}; ^{Ganahl et al., 2003}) considering only sex as a moderator variable and explaining gender differences using Bandura' social learning theory, which states that these differences are basically acquired through observational learning and imitation. On the other hand, face-ism was also studied by including the social role of the photographed person (^{Dodd et al., 1989}; ^{Sparks & Fehlner, 1986}), especially in political campaigns (^{Konrath et al., 2012}; ^{Konrath & Schwartz, 2007}; ^{Mattan & Small, 2021}). It was done based on the gender social role theory (^{Carli & Eagly, 2001}; ^{Eagly & Steffen, 1984}) to explain why women's heads are depicted more in competitive and professional environments than are men's. Later on, ^{Smith and Cooley (2012)} and ^{Cooley and Smith (2013)} studied the phenomenon using self-created internet profile pictures suggesting that there are also significant SAF differences for men and women on the internet. Still, they excluded the social role because SAF was studied only on Facebook and not on other SNS where people play different social roles.

So, according to the results of this study, it seems that all those theories fall short by including only gender as a variable to predict SAF. It appears that age and the specific type of SNS also play a role. So, SAF on-line seems not to have a unique and straightforward theoretical explanation. The social learning theory and the gender social role theory do not explain why within each SNS, women barely change their facial prominence while men increase it considerably when ageing. One possible explanation would come from the objectification theory (^{Fredrickson & Roberts, 1997}), proposing that women in Western countries are sexually objectified. According to the results of this study, women could integrate body objectification at a very early age and sustain it over the years, depicting more of the body in their profile pictures than men. And this would happen even when women's self-objectification experiences can trigger negative psychological states such as a more negative mood, lower self-esteem and confidence (^{Fox, Vendemia, Smith, & Brehm, 2021}). Also, evolutionary psychology may explain that younger women tend to self-represent more of the body than both men and older women. It might be "due to an evolutionary tendency to represent fertility by displaying the waist-to-hip area - a strategy that, from an evolutionary perspective, increases the likelihood of producing offspring" (^{Read, Pavelko & Hwang, 2017}, p. 3). In the same vein, evolutionary psychology might explain differences among different SNS. The increase of SAF on men and on professional SNS might be because the facial prevalence is an indicator of dominance, a dimension associated with status (Zuckerman, 1986, as cited in Read et al., 2017); then, when older the men in a professional SNS, the biggest is the SAF. However, none of them explains why the youngest men and women have similar SAF values. It could indicate that the new generations are represented more equally, or perhaps the very phenomenon begins at an older age.

The more variables are included to study face-ism, the more complex seems to be the phenomenon. In this line, it might explain the not significant differences found in previous studies, focused on only one SNS or in a narrower range of age (mainly, young people) (for instance, ^{Hum et al., 2011} and ^{Sczesny & Kaufmann, 2018}, focused on young people on Facebook and professional networks, respectively). The sex-gender construction is developed across biological, psychological, and social dimensions and is transformed through life experiences and learning. Probably, as a gender issue, it has complex theoretical frameworks integrated with multiple causes.

Practical implications

The present findings might have different practical implications. First of all, they might help plan and develop intervention programs, addressed mainly (but not only) to girls who feel the pressure of self-objectification in SNS to be accepted by their peers in particular, and society in general, by self-depicting their body. This pressure is perceived over the years in the case of women, regardless of the SNS used. Psychologically empowering women will allow them to behave beyond gender stereotypes which form part of their own gender identity. Several training programs are developed at this moment in Spain that focused on psychological (^{Machín-Rincón & Cifre-Gallego, 2020}) and economic empowerment (i.e., ^{Laguna-Sánchez, Segovia-Pérez, de la Fuente-Cabrera, & Vargas-Pérez, 2021};) which might help them to get rid of it.

Focusing on age, our results show that young people (both boys and girls, although mainly girls) self-depict their body in social but mostly dating SNS. Even boys manipulate more their body pictures than girls (^{Gioia et al., 2021}). Also, some studies show that older adolescents and girls are more likely to post images of themselves on another social SNS, Instagram, associated with peer norms (^{Van Ouytsel, Walrave, Ojeda, Del Rey, & Ponnet, 2020}). Then, as proposed by ^{Gioia et al. (2021)}, media literacy programs might be valuable tools to educate teenagers about their own actual body image, about culturally and peer-to-peer promoted body standards, and about their selfie-sharing on social media.

Also, results might have significant consequences in private and professional life, as it happens when searching for a job. In this case, older job seekers are less versed in on-line the self-presentations on professional SNS (i.e., LinkedIn) (^{Krings, Gioaba, Kaufmann, Sczesny, & Zebrowitz, 2021}). Then, avoiding gender behaviours (i.e., not increasing face-ism when getting older in the case of men) might help them undo gender x age stereotypes that could facilitate their incorporation into the labour market.

Summing up, our results highlight the potential of gender stereotypes in our self-depiction on SNS, regardless of the type (dating, social, professional). We are aware that sometimes they are used strategically to take advantage of their image, as politicians do (either emphasizing or rescinding gender-based stereotypes) (^{Mattan & Small, 2021}). However, from our point of view, society should work to undo those gender stereotypes to provide a more equitable and fair society for everybody regardless of their sex.

Limitations and future trends

Our study's main limitation lies in being cross-sectional and does not allow claims on causality. So longitudinal studies are needed to examine the causal dynamics. For example: will this growth in SAF depend on the sex × age interaction change over time in a longitudinal study? SAF differences may tend to disappear over time for younger generations, and therefore, SAF would be a good sexism indicator? Future research will have to elucidate this aspect. However, this study has a strength considering three important variables for the face-ism phenomenon at time: age, sex, and SNS, in addition to the interaction of age and sex. If each variable had been studied separately, the obtained parameters (coefficients, probabilities, errors, etc.) would not have evidenced real values, producing unreliable statistical coefficients (^{Gujarati, Porter & Gunasekar, 2013}).

Today, a profile picture on Facebook offers a significant first attempt to construct one's on-line identity (^{Hum et al., 2011}), but also to create their private (Badoo) and professional (LinkedIn) identities. SNS are becoming new virtual places where traditional body ideals and old gender stereotypes are reproduced, perpetuated, and reinforced. Given the media's role in influencing of women's self-image (^{Ghosh, 2005}) and young women with an appearance comparison tendency reporting negative mood after exposure to SNS images (^{Fardouly, Diedrich, Vartanian & Halliwell, 2015}), more intersectional research is needed to understand better the impact of gender on social representations and health. It would mean including more variables like social status, personality traits (i.e., competitiveness and narcissism; ^{Martin, 2020}), education and cultural levels, nationality, profession, ideology and culture (^{Nguyen, McDonald, Nguyen, & McCauley, 2020}) and, especially in adolescents, peer norms (^{Van Ouytsel et al, 2020}). Only by understanding this phenomenon and a whole, and the underlying mechanisms that support it, an equal future world is possible.

Data availability statement

The data that support the findings of this study are openly available in Universitat Jaume I repository: http://repositori.uji.es/xmlui/handle/10234/192402