You are on page 1of 77

Gender Presentation and Feedback on Instagram

by

Katherine O’Doherty

Presented in Partial Fulfillment of the

Requirements of Independent Study Thesis Research

Supervised by

Dr. Clayton

Department of Psychology

2021
Gender and Instagram 2

Table of Contents

Acknowledgements……………………………………………………………………………..…3

Abstract…………………………………………………………………………………………....4

Introduction………………………………………………………………………………………..5

Foundations of Gender Roles……………………………………………………………..5

Traditional Media………………………………………………………………………...11

Instagram…………………………………………………………………………………19

Gender and Instagram……………………………………………………………………23

Intersectionality…………………………………………………………………………..28

The Current Study………………………………………………………………………………..33

Methods…………………………………………………………………………………………..35

Results………………………………………………………………………………………...….47

Discussion………………………………………………………………………………………..59

References………………………………………………………………………………………..66

Appendix…………………………………………………………………………………………70

List of Tables

Description of Variables………………………………………………………………………..37

Gender of Subjects in Photos…………………………………………………………………….47


Gender and Instagram 3

Race and Ethnicity of Subjects in Photos………………………………………………………..47

Number of Comments on Posts………………………………………………...………………..47

Number of Likes on Posts………………………………………………………………………..48

Sponsored and Non-Sponsored Posts……………………………………………..……………..48

Setting of Posts…………………………………………………………………………………..49

Gendered Posing……………………………………..…………………………………………..49

Types of First Comments…………………..……………………………………………………50

Types of Second Comments………………………………………………………...…………...50

Types of Third Comments ….…………………………………………………………………..51

Types of Fourth Comments …………………………………………………………………….51

Types of Fifth Comments ……………………………………………………..………………..51

Gender of the Commenter……………………………………………………………………….52


Gender and Instagram 4

Acknowledgements

I want to thank my parents for their support during college and especially during the

writing of my Independent Study. I want to thank Dr. Clayton for her guidance all throughout the

process of my Independent Study. I also want to thank my friends for being there for me. I

especially want to thank Morgan, Taylor, Peyton, and Andrew. I couldn’t imagine college

without you.
Gender and Instagram 5

Abstract

Gender is such an important identity that children are taught from a young age how to

present themselves according to their gender. Social media has provided a new way to present

identity and the self to others, and while this has been looked at in previous studies, the feedback

that different presentation receive hasn’t been as studied. In this study, I looked at 214 of the top

Instagram posts and the first 5 comments under each post to see how people present gender and

how users respond to this presentation within different Instagram communities while taking into

account the intersection of other identities like race and ethnicity in the presentation of gender

online. I found that overall people conformed to traditional gender roles, though white people

tended to conform more to traditional gender presentation. I also found that non-sponsored posts

adhered to the idea of masculine and feminine spheres more than sponsored posts and posts from

company accounts. Additionally, I found that number of comments gave more significant

information about engagement with popular posts than number of likes did, suggesting that ways

of measuring engagement might need to be examined. In many ways, Instagram has become

another way of upholding, performing, and enforcing traditional gender roles.


Gender and Instagram 6

Introduction

Gender is one of the most important and profound way to self-identify, and through it an

individual often builds a sense of self through personal judgements of the self in terms of

masculinity and femininity (Goffman, 1977). Every society has its own ideas of what is

masculine and feminine and what is essential to a gender. In the United States and many other

western countries, women are idealized as pure, fragile, valuable, and as the givers as well as the

receivers of love and care. Men also tend to express that women need to be protected from the

“harsher things of life” and continue to idealize women through motherhood, innocence,

gentleness, and sexual attractiveness (Goffman, 1977, p.308). This places them at a disadvantage

and brings up the question of the consequences for women who do not fit these idealized

standards.

Foundations of gender roles. Sociologist Erving Goffman did a lot of early research on

gender, gender presentation, and the media that is relevant to psychology today. Goffman

believed that gender is a “learned, diffuse, role behavior” that social interactions and social

structures are built on in modern industrial society (Goffman, 1977, p. 301). At birth, people are

assigned a gender which is confirmed throughout their lives and determines how they are

socialized. The label of male or female leads to different treatment, acquiring different

experiences, and different expectations from others. There are specific ways of appearing, acting,

and feeling associated with each gender which have been naturalized leading to the belief in

modern society that the divisions between men and women are due to biological differences.

While there are biological differences between male and female, there is no reason why this

should have an effect on modern industrial society and is not innately the cause of societal and

social differences. These differences can be seen even in subtle ways, like the literary biases that
Gender and Instagram 7

use “he” as a default, though this is slowly being replaced by “they”, and “man” as a stand in for

“human” or “humankind.”

The household, and specifically the nuclear family, is where children are first socialized

and engendered. Girls are taught to take a domestic, supporting tole while boys are given more

options within a competitive role (Goffman, 1977). The naturalizing of gender differences is

often very overt in the home. For example, boys might get harsher punishments due to the

conceptualization of girls as fragile. This socialization is often done in front of the opposite

gender in the home, so girls and boys can compare how they are treated which creates a greater

understanding of gender differences. In many ways, children define themselves and their gender

by their differences from each other, including this difference in gender. Overall, the

naturalization of gender creates a rigid divide between men and women often completely

separate from biological facts, though it is believed by those with a rigid binary view of gender

that these definitions are rooted in biology, and it leads to men and women scanning any ongoing

social activity for means through which to express gender.

The organization of society in the United States reinforces gender roles. The nuclear

family is the basic unit and is embedded in a mostly homogenous community with women

equally distributed throughout the households due to their positions as children and wives

(Goffman, 1977). These women are cut off from each other by the nuclear family and its

structure as well as their individual stakes in the nuclear family, since women have a father or

husband, which may prevent them from rebelling against that structure. The social roles of men

and women are created in a way that gives women less power and less rank by keeping their

lives centered around the household and restricting their use of public space. This is done
Gender and Instagram 8

through things like being kept out of religious and public office or being excluded from war and

hunting.

While women still achieve things, like careers or leadership positions, this is usually

under some sort of arrangement. The concern is more about under what arrangement they get

those things and not whether they are getting less overall as the arrangements may restrict their

access or require them to fit a certain mold of femininity (Goffman, 1977). For example, in

traditionally feminine jobs that require work with the public like flight attendants, attractiveness

is used in the selection of employees. The hiring process for men and women makes it so that

men are often in the presence of women that are both young and attractive. This makes it so that

men exist in a “socially constructed world having what appears to be all-male settings

strategically stocked with attractive females” (Goffman, 1977, p. 318) outside of the household.

Overall, men and women are kept separate by their gender roles in a way that causes them to

relate very differently to public life, especially since its contingencies are greater for women than

men.

Hawke revisited ideas about gender and gender roles, especially within the home and

family dynamics, in the early 2000s. Hawke claims that the structure of the American economy

is to blame for most aspects of gender roles in the United States (Hawke, 2007). The American

economy prior to World War II was a family-consumer economy that segregated the roles of

men and women through their roles of mothers and fathers. Men went outside of the home for

work and then to risk their lives in war, while women stayed in the home to take care of the

house and the children. This made the ideal family one with a breadwinning father, a mother who

maintained order in the home, and submissive children. This seemed very similar to what

Goffman discussed with his ideas of gender and gender roles.


Gender and Instagram 9

The main focus of Hawke (2007) was on the impact of feminism and feminist theory on

gender roles, which created many of the current differences from Goffman’s ideas of gender.

This change in the home and family dynamics started around World War II when women took

over jobs while men were at war (Hawke, 2007). Afterwards, everything was expected to go

“back to normal” with men at work and women in the home, leaving many women unsatisfied.

Then, the 60s and 70s brought a social rebellion against tradition, including traditional gender

roles, which brought with it new feminist ideas. Some of these ideas raised issues of diversity of

family structure, the relinquishment of personal freedoms, and the exploitation of immigrant

women as caretakers in families where both parents work. These ideas also included men and

women sharing household tasks and childcare while both of them had careers, something that is

increasingly becoming the norm.

In the last 30 years, there have been changes in the gender equality attitudes of men and

women, showing that gender roles or adherence to those gender roles may be changing as well

(Hawke, 2007). The role of the husband and father within a family has become more complex to

allow him more caretaking and household duties. The balance of family and career is now an

issue for both mothers and fathers, as both roles allow for careers and caretaking. Overall, there

is more distribution of work within the household and the family while there is also more

flexibility within the roles of the family.

These changes have led to less strict definitions of what it means to be a mother or a

father which has led to changing ideas about what it means to be a man or a woman, since

definitions of gender tend to come from the family structure in the United States (Hawke, 2007).

Gender roles are currently less clearly defined, especially within a marriage or a household. This

allows individuals and families to make the best decisions for them with this increased freedom.
Gender and Instagram 10

Still, these gender roles are subconsciously or consciously upheld in individuals and groups that

are slow to change their ideas.

The workplace and media, whose standards of presentation are often reflective of

traditional gender roles, are often slower to pick up on progressive ideas. This means that

workers may feel they have to adhere to gender roles in order to get or keep a job. This can be

seen in formal and blatant rules, like dress codes, and in subtle social rules, like how mannerisms

from men and women are interpreted.

One of these mannerisms, smiling, is seen as a signal of motives and intention. Smiling is

often used to help an individual reach their own goals by using the expectations of others (Vrugt

and Van Eechoud, 2002). In a study by Vrugt and Van Eechoud , part-time jobs were advertised

to students on the university bulletin board. Jobs were then classified as either masculine or

feminine and high status or low status by 10 independent judges who read the job descriptions.

Four groups of fifty judges also rated which jobs emphasized the importance of social contacts.

90 students had their photos taken to see how they would portray themselves to make a good

impression for the given job. The study found that women smiled more only when social

contacts were important to a job, which is consistent with gender-role expectations and brought

about by those gender-role expectations. Women are expected to focus more on social relations

than men, so jobs assumed to be more appropriate for women heavily involve social contacts

while social contacts aren’t as important for jobs deemed appropriate for men. While men’s

smiling behavior was also according to the expectations of others, it was found that women had a

better image of the expectations of those they were interacting with.

Both men and women displayed and exaggerated gender stereotypical traits associated

with a gender stereotypical job with all jobs except for feminine high-status jobs where gender
Gender and Instagram 11

stereotypical traits weren’t seen as necessary to make an impression (Vrugt and Van Eechoud,

2002). This could be due to high status jobs not usually being included in stereotypical

femininity. Additionally, low status jobs considered either masculine or feminine were all in the

service sector while gender neutral ones were not.

The Bem Sex Role Inventory, or BMRI, measures masculine and feminine traits to

determine sex role identity (Gaa et al, 1979). Masculinity and feminine are separate measures

instead of two extremes on one continuum. If someone scores above the median on both

masculine and feminine scales, they are considered androgynous. If they are above the median

on the masculine scale but below the median on the feminine scale, they are considered

masculine. If they are above the median on the feminine scale and below the median on the

masculine scale, they are considered feminine. If they are below the median on both, they are

considered undifferentiated.

Masculine traits include self-reliant, defends own beliefs, independent, athletic, assertive,

strong personality, forceful, analytical, has leadership abilities, willing to take risks, makes

decisions easily, self-sufficient, dominant, masculine, willing to take a stand, aggressive, acts as

a leader, individualistic, competitive, and ambitious (Puglisi, 1980). Many of these traits are

associated with aggression as well as traits that are helpful in the workplace, further asserting

that men are meant to be in traditionally masculine realms, like the workplace.

Feminine traits include yielding, cheerful, shy, affectionate, flatterable, loyal, feminine,

sympathetic, sensitive to needs of others, understanding, compassionate, eager to soothe hurt

feelings, soft-spoken, warm, tender, gullible, childlike, does not use harsh language, loves

children, and gentle (Puglisi, 1980). Many of these traits, like loves children and sensitive to

needs of others, already assign women the mother and caretaker role. Some traits, like gullible
Gender and Instagram 12

and childlike, also infantilize women. This confines women to their traditional role in the home

by portraying them as fragile and caring, so they both are well-suited for the traditionally

feminine realm of the home and may be unsafe outside of it.

A study of the BSRI compared the masculine and feminine subscales to each other and

then compared the traits within the subscales (Puglisi, 1980). The initial study found that on

average, the masculine subscale was more socially desirable than the femininity subscale.

However, it was found that childlike, gullible, and shy were rated so undesirable that it might

have caused the difference between the subscales. Another study took out these traits and

replaced them with three more desirable feminine traits-charming, graceful, and gracious-which

were chosen by asking different participants what traits the most feminine person would have.

After the traits were replaced, there was no longer a significant difference between the social

desirability of the masculine and feminine subscales (Puglisi, 1980). This means that studies

using these measures of masculinity and femininity may be biased since they include the original

less desirable feminine traits that participants might be more hesitant to identify with, and it

raises questions about the ways that we measure and characterize masculinity and femininity.

Traditional Media. The traits and roles expected of men and women in daily life are

often reflected in traditional media. The media often reflects the most common or

uncontroversial attitudes of the general public to appeal to as many people as possible, so the

media tends to portray men and women in ways that the general public would agree with. The

media also tends to exaggerate many things to grab attention or to get a certain point across. This

could lead to the exaggeration of traditional gender roles in things like television,

advertisements, and magazines.


Gender and Instagram 13

Goffman was also interested in media and marketing. He was particularly interested in

the new forms of construction of self in a society “driven by consumption, marketing, and

media” like the United States and many countries currently characterized by advanced capitalism

and commercial realism (Hancock and Garner, 2015, p.163). This is even more relevant today

than when Goffman was writing. He looked at new forms of socialization and enculturation of

individuals. He was especially interested in the ways that advertisements reproduce gender,

specifically through framing practices. These advertisements convey conventional and abstract

representations of gender. Gender representations in advertisements are meant to reflect the

gender displays that would typically be seen in daily life rather than fully and directly represent

all gendered behavior.

The way individuals understand gender is always distorted and incomplete (Hancock and

Garner, 2015). Advertisements conflate culture with real daily life, meaning the way individuals

see and understand the world follows the formulas and structures presented by mass media,

because advertisements provide context for the experiences of individuals that consume them.

Especially in the age of social media, interaction that is mass mediated is so intertwined with

personal interactions to the point that there is almost no distinction between the two. Celebrities,

models, and influencers become the standard of beauty that the average person compares

themselves to. In recent decades, this has often been considered a major factor in rises in eating

disorders and body image issues.

By trying to conform to idealized gender norms, people present themselves in ways that

exaggerate gendered behaviors though they can never live up to these idealized gender norms

(Hancock and Garner, 2015). This is even more important with the relatively new use of

photoshop as well as traditional makeup, hair, and lighting. Digital editing makes the norms
Gender and Instagram 14

created by mass media even more impossible for the average person to follow. Additionally,

mass media emulates whatever styles, tastes, and images are popular in order to appeal to the

largest amount of people, and then the masses emulate styles, tastes, and images they see in mass

media, creating a looping effect where the media copies the masses which copy the media. This

then further blurs the lines between media and real life which weakens the autonomy and

authenticity of individuals. However, Goffman wasn’t entirely pessimistic about the self,

believing there were some stable characteristics of self that aren’t as susceptible to social and

media influence. This means he didn’t think the self was ever completely absorbed or dissolved,

so while the media is still hugely influential and at times destructive for the self, the authentic

self still remains in one way or another.

Advertisements are a constant part of modern life and are often more difficult for

individuals to control exposure to. Advertisements are meant to simulate everyday life for the

average viewer, so they can see the products or services in the advertisements being

implemented in their own lives (Furnham and Skae, 1997). This includes the people, social

dynamics, problems presented, and settings. This can be seen in things like infomercials which

often present a common problem that their product can fix in a common setting, like the home.

Since advertisements are usually set in a setting that reflects everyday life, they can send us

messages about what our daily life should be like and change how we view the world around us.

One way advertisements can do this is through representation-both in amount of

representation and how social groups are shown in these representations. One study looked at

British television advertisements on the most widely watched British television channel every

day for a week at multiple times throughout the day (Furnham and Skae, 1997). Men and women

were shown to differ in their portrayals by credibility, location, roles, ages, and products.
Gender and Instagram 15

Women were shown with body products and “other products,” like children’s toys, and men

were shown with auto products and sport products. Women were also more likely to be shown

with products that are social or self-enhancing, and men were seen more with products that were

pleasurable. The “rewards” of these products, or what the public is told comes with having this

product such as increased status with a nicer vehicle, seem to follow traditional gender roles and

stereotypes where women are associated with sociability and men are usually seen as seeking

pleasure, especially sexual pleasure since aggressive heterosexuality is often seen as masculine.

In regard to speaking parts, men were more likely to give the end comment, even if a

woman had been speaking throughout the advertisement (Furnham and Skae, 1997). Men and

women were equally likely to give an argument and were equally likely to give either a fact or

opinion as the argument. Men were more likely to do voice-overs, but women were just as likely

to be visually portrayed when they did do a voiceover. There was no gendered effect on who was

seen as an authority on a product, though there was a difference in what products men and

women were shown with.

The appearance of men and women were also very important, especially when the

advertisement is in a visual medium like television. Women were overrepresented in the under

30 category and underrepresented in older age categories, which may be a reflection of society’s

expectations that women should fear aging and be valued for the beauty (Furnham and Skae,

1997). Men and women were equally represented in all locations, but women were seen in the

home more, though this wasn’t a significant finding. Additionally, women were in dependent

roles more often than men, and men were more likely to be in a narrator or interviewer role. This

likely reflects the stereotype of men as more knowledgeable and independent while also

reflecting the higher status men tend to have in society.


Gender and Instagram 16

A study of four different countries-Malaysia, Japan, Taiwan, and the United States-

looked at changing gender roles in prime-time commercials as a reflection of changing gender

roles in those countries (Bresnahan et al, 2001). Men and women were in an equal number of

advertisements, so the way they’re represented was more important in this study. The depictions

of men and women were important in advertisements, because ads are constructed to reinforce

the “image of sex most familiar to and comfortable for their target audience” (Bresnahan et al,

2001, p. 129). Across all four countries, product endorsement conformed to gender stereotypes.

For example, in America women promoted drugs and real estate significantly more than men.

Again, men narrated more than women.

Women were more likely to be shown in the home than men, and men were more likely

to be outside the home in Japan, Malaysia, and the United States (Bresnahan et al, 2001). Men

were depicted significantly more than women in the workplace, which was the only setting in

which men were seen more. Greater professional visibility of women is seen specifically as a

marker of freedom from traditional women’s roles, because the workplace wasn’t traditionally

for women. In Japan, women were shown in traditionally feminine roles 33% of the time while

men were shown in traditionally masculine roles 20% of the time, though non-stereotypical roles

were the predominant gender roles shown in prime-time commercials. In Taiwan and Malaysia,

they were shown in non-stereotypical gender roles in half of the commercials. In the United

States, non-stereotypical gender role presentation was only about 29.5% of role depictions.

Additionally, role reversal was often used in United States’ commercials to address stereotypes

by showing women in traditionally masculine roles.


Gender and Instagram 17

Overall, this study showed that presenting gender roles in prime time commercial is still

seen as profitable in these four countries, but there is a changing status of men and women

(Bresnahan et al, 2001).

A study by Furham and Bitar looked at British advertisements and found that men were

more likely to promote products in advertisements (Furnham and Bitar, 1993). Women were

younger than men, which rarely changed over time. Men were usually associated with food

products, especially those popular with men in Britain like beer, and with nonmaterial products

like banking or insurance. Women were shown with food products and body products.

Additionally, the way men and women were seen in these advertisements were different.

Over half of male central figures were seen as authorities while only 39% of female central

figures were seen as authorities (Furnham and Bitar, 1993). Men were likely to be interviewers,

narrators, and celebrities in these advertisements and were shown in the workplace or

unspecified locations. Women were usually shown as dependent on others and shown in the

home. Again, men were more likely to be voice-overs while women were more likely to be

presented visually, and 75% of men had end comments in contrast to 33% of women.

Television can be just as important to concepts of self as advertisements with the main

difference being that there is more control and choice in television. Two and a Half Men, a

popular sitcom that consistently held a prime-time slot on broadcast television from 2003 to

2015, is a good example of the typical popular television show (Bradley, 2013). It presented

heterosexuality as the norm and as intrinsically tied to normative gender presentations. Non-

normative gender characteristics were consistently pointed out to the viewer and connected to

homosexuality. It explicitly associated masculinity in females with homosexuality while

constantly reasserting the stereotypical ideas that men are promiscuous, and women should be
Gender and Instagram 18

attracted to typical male behavior. This show perpetuated the idea that men should be sexually

aggressive and that if a woman were to seek sexual pleasure, it would emasculate the man.

Women were also seen as failing in a traditionally feminine role when they stepped

outside of traditional gender characteristics, so a working woman would either be seen as failing

to be a mother or unmarried (Bradley, 2013). Men were much more shamed by other characters

for transgressions and missteps in regard to gender and sexuality; in fact, they were 50% more

likely to be told to act properly and were considerably more likely to be dehumanized. Both men

and women were reprimanded for gender missteps, but this was done in different ways and to

varying degrees.

In a study by Bazzini et al, Women’s Health and Men’s Health magazines were both just

as likely to objectify body parts in captions, but how they were objectified did differ (Bazzini et

al, 2015). Men’s Health captions usually were about a muscular body while Women’s Health

captions were usually about thinness. This was likely related to the differences in body and

beauty ideals for men and women. Additionally, Men’s Health captions were simultaneously

concerned with “bulking up” and “slimming down,” bringing up concerns for the conflicting

messages. Both men and women’s captions were concerned with weight loss, which was

assumed to be due to the United States concerns with obesity and the stigma surrounding being

overweight.

None of the captions were actually concerned with health for both magazines and were

instead focused on conforming to masculine or feminine beauty ideals with body improvements

being framed solely as for aesthetics (Bazzini et al, 2015). Secondary captions that were

concerned with dieting, weight loss, appearance, and exercising for the sake of appearance were

more common in Women’s Health magazines, and Women’s Health magazines promoted these
Gender and Instagram 19

as goal-oriented behaviors that will lead to beauty and thinness. Men’s Health messages were

more about physical strength, muscularity, and performance than appearances, so they were

encouraged to exercise for the sake of exercising, though again the health purposes and benefits

weren’t emphasized.

While captions may have been equally objectifying, Women’s Health more frequently

depicted partially clothed individuals regardless of season, and overall, 52% of Women’s Health

cover models were partially clothed, portraying these women in a way that was more sexualized

than the men on Men’s Health magazine covers (Bazzini et al, 2015). It was believed that this

may convey the message that healthy activity is a goal-oriented behavior which will lead to

looking attractive and sexy.

A study by Plous and Neptune looking at 6 fashion magazines popular among different

demographics to compare the advertisements. They looked for racial and gender representation,

body exposure, posing, and clothing in the advertisements to examine both the amount of

representation and the type of representation that Black women, Black men, White women, and

White men receive in media (Plous and Neptune, 1997). Magazine advertisements contain

significant racial and gender biases which often effect Black women more than other groups.

Black women have seen the most changes in representation and types of representation in

magazine, but this is often due to starting in a worse place, making their representation seem

more equal in comparison. Concerning amount of representation, White people were seen more

in magazines with a predominantly White readership and Black people were seen more in

magazines with a predominantly Black readership. Interestingly, there were very few interracial

couples in all of the magazines.


Gender and Instagram 20

The clothing and posing of the people in magazine advertisements often speaks to gender

and racial stereotypes. This can be seen in the exposure of subjects in the advertisements.

Advertisements with women were four times as likely to include bodily exposure in comparison

to advertisements with men (Plous and Neptune, 1997). While the current level of bodily

exposure is the same for Black women and White women in advertisements, this is not true over

time. The bodily exposure of Black women started much higher in advertising while the

exposure of White women in advertisements has increased over time. This might be due to Black

women historically being hypersexualized, so they might’ve in the past been shown as more

exposed than what was seen as acceptable for a magazine. Additionally, the exposure of Black

women was more likely in magazines marketed towards Black women while the exposure of

White women was more likely in magazines marketed towards White women.

The amount of exposure seen in magazine models isn’t reflective of the amount of body

exposure seen in everyday life, especially in women (Plous and Neptune, 1997). In magazine

advertisements, all women were more likely to be seen in sexual attire, like underwear, than

men. Black women again were more sexualized through clothing in magazines. They were the

most likely to be seen wearing animal prints, and the majority of advertisements that included

Black women had them wearing animal prints. These advertisements were shown to be leaning

into the sexualization of animal prints through provocative posing, body exposure, and the

captions. The only depiction in advertisements that didn’t disproportionally affect Black women

was the symbolic lowering of women. The symbolic lowering of women where women are

depicted as physically lower than men was seen to apply more to White women than Black

women. This might be due to some sort of implied relationship in that posing and the lack of

interracial couples in advertisements.


Gender and Instagram 21

Instagram. Instagram is a photo-sharing network owned by Facebook. It exists as a

website but is primarily used as an app which is the second most downloaded free app in the app

store (Suciu). It’s the third most popular social media network and has more than a billion

monthly active users (Suciu). Part of Instagram’s popularity is due to how public the app is.

Anyone can like, comment, share, or save any post as long as it comes from a public account, as

opposed to a private account. The posts from private accounts can only be seen by those

following that account, and in order to follow the account, users need to send a request to follow

which then must be approved by whoever runs the account. This allows users to have the option

to post things just for those they accept as followers as well as the option to build a large public

following to share their photos with everyone on the app.

Instagram allows users to post up to 10 photos in one post and even has some basic photo

editing functions, such as the ability to adjust the contrast, lighting, and orientation of the photos.

This allows users to create the best-looking photos possible without having their own photo-

editing software which will be more appealing to other users who come across their photos,

allowing them to get more positive feedback and possibly grow their following. Users can add a

caption to their post which often include hashtags that allow for the categorization of the post,

such as the #selfie that usually is seen with pictures a user takes of themselves. Users can also

“tag” others in a post using the @ feature in the caption or comment, which is done by typing a

username directly after the @ symbol which will notify that user of the post and allow others to

click on the tag and go to the profile of the tagged user. This allows users to show who is in the

photo and increase the likelihood of people following them from that post. Usually, the users

would make similar posts and have similar interests, so this can help build an online community.

Tagging can also be done by tagging a user in the image through the “tag people” option that
Gender and Instagram 22

allows users to click anywhere on their own photo and type a username. This works similarly to

the other tag function but can only be done by the owner of the post and pops up on the image

itself after the image has been posted.

Hashtags are specifically important as they allow users to categorize their posts so that

more people interested in that hashtag see it, directing more users to their post and allowing them

to connect with users that have similar interests. Online communities and niches tend to use the

same or similar hashtags. Hashtags allow people to find online communities and build the

communities that they’re a part of. Hashtags have their own pages that can be reached through

either clicking on the hashtag or searching for it in the search bar on the explore page. The page

of any hashtag includes two different feeds-the popular and the recent feeds. The popular feed

shows the most interacted with posts that include the hashtag in the captions or comments of the

post. The recent feed shows the most recent posts that include the hashtag in the captions or

comments of the post. The hashtags can be followed just like accounts, making them show up on

a user’s home page and affecting their explore page. The explore page is a feed of posts from

accounts the user doesn’t follow. It is generated by an algorithm that takes into account the

categories of posts and accounts that the user interacts with-included likes, comments, shares,

and bookmarks. It also takes into account how many other users have interacted with posts and

accounts to show the user the most interacted with posts that are within their interests. A similar

algorithm decides what order comments on photos are shown in. Comments with more likes,

comments made by “verified” accounts, comments made by the original poster, and comments

with a lot of replies tend to be shown before other comments. This means that the first comments

usually have more interactions and are therefore more indicative of the general response to the

post. The explore page allows users to see posts related to their interests by people they aren’t
Gender and Instagram 23

following which allows them to either be introduced to new communities or to be introduced to

other users who are a part of the same communities as them. The home page of the app or

website is a feed of photos from the users an account follows that is ordered according to an

algorithm that takes into account how often the user interacts with that account and how much

other users have interacted with that post.

Instagram has a fairly diverse userbase, though this isn’t necessarily true for age. 64% of

Instagram’s users are 18-29 years old, meaning that 2 out of every 3 adults in that age group has

an Instagram account (Suciu). This is likely due to Instagram being a newer social media site,

unlike Facebook which has a reputation for being primarily used by older adults. This could also

be due to Instagram being less easily accessible via website to the point that new posts can only

be made through the app. Otherwise, Instagram has a diverse userbase.

Instagram is used almost equally by men and women with males accounting for 51% if

Instagram’s total users (Suciu). Additionally, Instagram’s use is not confined to the United

States. Only 11% of Instagram’s users are in the United States, which accounts for 112.5 million

people (Suciu). 70 million active users are in Brazil, 69 million active users are in India, and 40

million active users are in Russia (Suciu). This means that Instagram’s influence is wide

reaching but also that it is a platform that is adaptable enough to be appealing to people from

many different cultures and with many different interests.

Unlike other social media platforms-like Facebook, Snapchat, and YouTube, people use

Instagram both to see the people that they know and influencers (Dhillon). This makes Instagram

an ideal social media platform to study, because users are seeing their friends and family, social

media influencers, and traditional celebrities all in the same place, often putting the posts of “real

people” in competition with those of influencers and celebrities. This also makes posts from
Gender and Instagram 24

celebrities and influencers seem more real, especially with the introduction of Instagram stories

and IGTV. Users “empathize with behind-the-scenes, fly-on-the-wall voyeuristic viewing of

influencers’ insta-stories,” (Dhillon) ultimately making the one-way relationship users have with

influencers seem more like a friendship, especially since many address their followers as though

they are friends. The introduction of Instagram stories and IGTV also encouraged users to spend

more time on Instagram and encourages influencers to use Instagram as their main platform

while allowing Instagram to push more ads that at first glance might appear to be just another

post in the Instagram feed (Dhillon).

Instagram’s advertisements don’t stop with the companies that pay Instagram to promote

them. Companies often directly pay influencers to post photos with their products or give them a

shout out on their insta-story. These posts became so popular that there are now 14.6 million

posts on Instagram that use the hashtag #ad. This doesn’t account for the posts that try to hide the

fact that they’re sponsored or used different disclaimers, further blurring the line between

sponsored and non-sponsored content or ads and non-ads. Additionally, businesses and brands

create their own Instagram accounts to promote themselves. This has proven to work well since

80% of accounts on Instagram follow at least one business or brand, and 200 million accounts

visit business profiles every day (Suciu). This has led to Instagram having a reputation of being

good for brand marketing and brand awareness (Business Insider India). In fact, over the past

two years companies have increased their spending on Instagram ads, and this has increased even

more in 2020 than in previous years (Business Insider India). This is likely due to the wide use of

Instagram among young people across the globe.

Gender and Instagram. There are many different usages to social media from personal

entertainment to more professional usages. One example of these professional uses are celebrity
Gender and Instagram 25

social media profiles, including professional athletes. Athletes are able to use social media to

their advantage by having complete control over how they’re portrayed on their social media

accounts, like Instagram (Smith and Sanderson, 2015). Often, they use social media to counteract

unfavorable portrayals in mass media as well as to engage in parasocial interactions with fans.

To a lesser degree, most individuals use their personal social media accounts to control their

image as well.

Online behaviors reflect the age, gender, emotion, interests, attitude, and personality of

users (Song et al, 2018). Hashtags can reflect a variety of things from image content to user

intentions that can create potential connections to other users and their characteristics. In a study

that looked at what different demographics were posting about, they looked at both differences in

age and in gender. They found that differences in topics of interest were more diverse between

gender groups than between age groups where topics more closely related to people were more

common in teens than adults. Teens were also more likely to post selfies than adults. These

topical differences might be due to the social development and changes that happens during

teenage years. Topics were more diverse for female users, who were more interested in fashion,

beauty, and decoration, than male users, who were more interested in design, cars, and urban

areas. Females were also more likely to post selfies than males, including both individual and

partner selfies. This is partially due to females exhibiting more social exhibitionism and

extraversion than males.

Social media allows users to curate their image and exert control over visual messages

they share with others through posing, composition, editing, and captioning (Butkowski et al,

2020). Often, users end up using the same stereotypical presentations seen in traditional media,

such as stereotypical gendered cues like those that depict women as disengaged and subordinate
Gender and Instagram 26

and men as dominant. In this study, young women completed a questionnaire about gender, race,

sexuality, and political ideology and had the 10 most recent selfies pulled from their Instagram

pages to analyze their gender display and the feedback they received. Coders looked at 6

different gender display variables and rated them on a scale of 0 to 6 for exaggeration of the

pose. They also looked at the number of likes, number of comments and whether the first five

comments were positive or negative.

Some traditional gendered cues include feminine touch, ritualization of subordination,

and licensed withdrawal. Feminine touch refers to when the subject of a photo uses their hands to

feature object or their body in a delicate and passive way. Licensed withdrawal refers to images

where subjects seem distracted or psychologically removed from their surroundings, including

looking away. Ritualization of subordination refers to a wide range of poses where subjects seem

yielding, unstable, or physically low to the ground. This includes canting, bending the neck or

body, low posture, being seated or lying down, and imbalance.

A study by Butkowski et al (2020) looked at a sample of young women’s selfies and

analyzed their content, looking for gender cues and engagement on posts. They found that gender

cues were used in most photos, but they were implemented in subtle and de-emphasized ways

where 95% of photos included at least one gendered cue but most were expressed with some

restraint compared to traditional media and advertisements (Butkowski et al, 2020). This could

be due to the belief that these gender expressions should be innate and effortless, so a de-

emphasized gender cue seems more natural, reflecting the way women are expected to present

themselves in a way that is both idealized and authentic. The most common poses were canting

and loss of control, and these poses were performed in the most exaggerated way. Imbalance and

instability were the least common and least exaggerated pose, though this may be due to how
Gender and Instagram 27

difficult it is to take a photo of themselves while posing like this. Feedback was significantly

related to gender posing categories, so participants who used gender display in their photos

received more positive and neutral comments.

People display gender expressions in their selfies through what is on their person, like the

clothes they’re wearing, as well as through expressions used in taking the selfies, like posture

and facial expression (Döring et al, 2016). In a study by Doring et al., gender stereotyping was

evaluated by looking for feminine touch, ritualization of subordination, licensed withdrawal,

body display, kissing pout, muscle presentation, and faceless portrayal in photos. Feminine touch

was whether the subject of a photo was touching their clothes or themselves or was tracing an

object with their hands. Ritualization of subordination was measured through posture and

imbalance. Licensed withdrawal was either withdrawing gaze or loss of control over emotions.

Body display measured whether the subject of a photo was wearing sparse clothing. Muscle

presentation measured whether a person posed to show off muscles. Faceless portrayal measured

any instance where a picture didn’t reveal the subject’s face. There were gender differences for

all the measures except for posture, and the biggest gender effects were for feminine touch and

imbalance. Females used kissing pout and faceless portrayal more, and these had the biggest

gender differences. Males used muscle presentation the most. In feminine touch, imbalance,

withdrawing gaze, and loss of control, social media posts were more gender stereotypical than

advertisements in magazines, showing that Instagram might reproduce gender stereotypes more

than magazines.

Social media often compels users to use predetermined ways of presenting themselves to

their audience of online followers (Baker and Walsh, 2018). This means that users copy the

trends of other users and can use similar lighting, posing, and phrasing in the captions that are
Gender and Instagram 28

significant to the culture of Instagram to convey an image of them as a certain kind of user. This

also means that users are unlikely to try new types of presentation or types of presentation that

are common outside of their internet communities. Another way users portray themselves as a

certain kind of user is through hashtags. Hashtags, which make their posts discoverable, allow

them to grow their following, communicate meaning and identity, and signify community

membership with others that use similar hashtags. Captions and comments tend to reference the

community itself as if users were imagining the other that they’re performing for.

Social media sites encourage their users to seek out recognition and approval through

likes, followers, and comments which can push their images into the position of high-status

images on the hashtag page (Baker and Walsh, 2018). The styles and forms of top images tend to

be very similar to other top images under the hashtag, showing a cohesiveness in the

communities that use these hashtags. These top images are emulated by others seeking similar

recognition and approval. Through this emulation of others, Instagram can inform an individual’s

gender identity.

Many of these ideas can be seen in the clean eating community on Instagram, which

caters to traditionally feminine interests but can still be used to present a traditionally masculine

image (Baker and Walsh, 2018). The clean eating community is an online community centered

on following a very strict diet that eliminates “unclean” foods while increasing consumption of

plant-based foods. “Unclean” foods include processed foods, refined grains and sugars, and

foods high in fat or salt. The main reason for following this diet is the believed health benefits

that come from eliminating these foods. The users in this community are largely White,

heterosexual, and middle-class which likely also has an effect on how they chose to present

themselves and seek a supportive community. The clean eating community uses hashtags and
Gender and Instagram 29

buzzwords to identify each other and highly emphasizes the body, suggesting it is a direct effect

of clean eating practices. A study by Baker and Walsh looked at the connection between diet,

gender, and advertising through examining photos and videos under the #cleaneating and

#eatclean hashtags. The researchers looked at the nine top posts under the hashtags every day

until they reached 144 total posts. They then coded whether the post included kissing pout, a

before and after image, food, a glamour shot, muscle presentation, an advertisement, or a nature

shot. They also coded for sex of the subject and the number of subjects in the photo or video.

The study found that glamour shots were about 13% of clean eating posts and showed this

emphasis of the body. The focal point of the image was an often exposed and highly sexualized

body, and these images mostly featured women who displayed an idealized form of femininity.

Muscle presentation was also common, seen in about 42% of posts. These mostly featured men

that portrayed ideals of hegemonic masculinity through displays of physical strength, athleticism,

and competitiveness. Their dietary choices were framed as rational to compensate for the

feminine association with clean eating with the association between reason and masculinity.

Muscle presentation was also used as a sign of merit and achievement with an emphasis on

motivation, determination, and choices. These examples show how health on Instagram is based

on societal expectations and performances while health is also used for the performance of

gender displays. These findings also show that social media reinforces hegemonic conceptions of

gender as well as the commercializing of gender presentation of an individual to gain acceptance

and approval by the community.

Intersectionality. Gender and gender presentation in media and social media has been

shown to be affected by race and ethnicity. Social media usage varies by many demographic

variables. The main variables that determine social media usage are age, number of internet
Gender and Instagram 30

access locations, owning a cellphone with internet connectivity, and whether someone was

transgender or cisgender (Park et al, 2018). People with the highest social media usage are

usually non-Hispanic white individuals aged 18-31 with access to internet on a cellphone and

some college education.

Traditional media has a huge impact on how people see themselves and present

themselves, because it’s seen by the largest number of people and has been present in many

people’s lives for as long as they can remember. Representations in traditional media may be

more equal by gender than they are by race. It’s been argued that long term exposure to

stereotypical and biased portrayals of different groups could distort the perceptions of these

groups (Rubie-Davies et al, 2013). This makes the types of representation just as important as the

amount of representation itself.

One study looked at a sample of advertisements in New Zealand (Rubie-Davies et al,

2013). This study looked for proportions of racial and ethnic groups in advertisements and

compared them to the proportion in the general population. They also looked at the role that

actors had, what types of advertisements they were in, and how positively actors were depicted.

They found that advertisements actors were not representative of the proportions of the

general population from census data where white actors were represented most frequently in

advertisements overall and most represented in every category of advertisement (Rubie-Davies et

al, 2013). Even when advertisements had actors from different racial and ethnic groups together,

the white actors were still portrayed as the main characters while minority groups had supportive

or background roles.

Additionally, minority groups were often portrayed in accordance with stereotypes

(Rubie-Davies et al, 2013). For example, while Asian actors were rarely included in
Gender and Instagram 31

advertisements, they were only proportionately portrayed in banking and finance advertisements

and were completely missing from advertisements about leisure and family-related products.

This stereotypical portrayal of minority groups can also be seen in portrayals of the Maori and

Pasifika people. They were seen in advertisements warning viewers about drinking and driving,

gambling, smoking, and invoking violence towards family members as well as advertisements

about needing to read to children. All of these portrayals send implicit messages to viewers that

Maori and Pasifika behave in negative ways that correspond to stereotypes. In advertisements

where the products and actors are seen positively, Maori and Pasifika people are rarely included.

This also supports stereotypes.

Representation of gender and race intersect in advertisements. In video game

advertisements, both minority and female characters appear significantly less than white and

male characters (Behm-Morawitz, 2017). 82% of the characters were white, and 80% of

characters were males where there were more males than females in every racial and ethnic

group. Male video game characters were seen more in violent video game advertisements than

female characters, which conforms to gender stereotypes, and there were so few minority female

characters that an intra-group comparison couldn’t be made, though there was no significant

difference of race and ethnicity alone in representation in violent video game advertisements.

Violent behavior, however, did show a significant relationship with race where Asian characters

were more likely to be violent while White characters were less likely to be violent. This was

mostly due to the male characters, since Asian male characters were more likely to be violent

than White male characters. Additionally, Asian characters were most likely to appear with

weapons. White males were the least likely to be seen with weapons, though this wasn’t a

significant finding.
Gender and Instagram 32

Attractiveness and sexualized portrayals of characters in video game advertisements also

varied by race and gender. White characters were disproportionately portrayed as highly

attractive while Black characters were depicted as the least attractive (Behm-Morawitz, 2017).

Additionally, female characters were portrayed as highly attractive significantly more than male

characters, which was contributed to most by White and Asian female characters showing an

effect of race on gender differences. Black characters were also depicted as “very muscular”

significantly more than White characters. There were intragroup differences of race and gender,

specifically with the muscularity of White characters, where White females were less likely to be

depicted as highly muscular while White males were most likely to be depicted as highly

muscular. White characters were also significantly more likely to be portrayed in provocative

attire than Black and Asian characters, and female characters were depicted in provocative attire

significantly more than male characters. White characters were significantly more likely to have

a sexualized appearance than Black or Asian characters, and female characters were more likely

to have a sexualized appearance than male characters. There was also an interaction of gender

and race in sexualized appearance where White females were more likely to be depicted with a

sexualized appearance than Black and Asian females. Female characters were also more likely to

have an unrealistically small waist, with an interaction of race so that White female characters

account for most gender differences in this category. Female characters were disproportionately

shown in states of undress and with a larger amount of skin showing than male characters.

Overall, the portrayal of video game characters followed gender and racial stereotypes with a

relationship between the two showing how identities can intersect and have more than just an

additive effect in media and in the lives of real people.


Gender and Instagram 33

These different depictions of race and gender and the relationship between the two in

media has an effect on real people and their real lives. The BEM Sex Role Inventory, or BSRI,

was studied to examine traditional western gender roles and how they hold up to popular current

conceptions of gender (Hershey, 1978). For a study by Hershey, participants were asked whether

certain activities which are commonly stereotyped by sex are appropriate for either sex,

somewhat more appropriate for one sex, or much more appropriate for one sex. This allowed for

the creation of a measure of how strongly respondents believed in strictly assigned gender roles

on the basis of sex. When comparing the answers of black and white respondents, it was found

that there were barely any racial differences in gender-role identities and in attitudes about

gender roles. Additionally, departure from traditional nuclear family structure, like having

unemployed or absent fathers, didn’t lead to flexible attitudes and identities in regard to gender

roles, and often it led to stricter adherence to gender roles.

Personal measures of masculinity and femininity have been found to have different

results depending on race and ethnicity of the individual. In a study of queer women, measures of

masculinity and femininity were compared in Black, White, and Latina women (Everett et al,

2019). Femininity was positively correlated with internalized homophobia, and Black and Latina

women were more likely to report internalized homophobia. They also saw most of the effect of

the negative association between femininity and internalized homophobia. Alternatively,

masculinity was associated with less internalized homophobia for Black women. Black and

Latina women were more likely to report victimization in general and were more likely to report

discriminatory experiences regardless of masculinity and femininity measures. Masculinity in

White women were positively correlated with victimization, but in Black and Latina women,

masculinity was negatively correlated with victimization. This could be due to masculinity
Gender and Instagram 34

making White women seem more visibly queer and therefore more likely to be discriminated

against and victimized while Black and Latina women are already discriminated against and

victimized for their race, which seems to be supported by self-reported discrimination.

Masculinity was negatively associated with self-reported discrimination for Black and Latina

women, but for White women masculinity scores increased self-reported discrimination.

Femininity was negatively associated with stigma consciousness for Black and Latina women.

There was no effect of femininity of stigma consciousness for White women, but masculinity

was positively associated with stigma consciousness for white women and negatively associated

with stigma consciousness for Black and Latina women. These effects of masculinity and

femininity were very different depending on race, showing how intersectional approaches are

very important to understanding gender.

When individuals have more control over how they are presented, gender presentation

may no longer be due to what is palatable to a large audience but instead is assumed to be what

the individuals wants to project about themselves to an external audience; however, exposure to

traditional media may change their internal understanding of their gender and their external

presentation of it as they seek to conform to the images they’ve seen all their lives (Kapidzic et

al, 2015). On social media, all choices about the photographs are made by the user who can use it

to create whatever self-presentation they’d like. When comparing the profile photos of teenagers,

there was a gendered difference in self-presentation with an interaction of race. In a study,

profile picture of 400 users on the most popular online teen chat websites were examined to see

how teens were presenting themselves online. They were coded for race, gender, whether or not

they had a profile photo, distance from the camera, dress, and posing behaviors. They found that

girls were closer to the camera, posed more seductively, and were dressed in more revealing
Gender and Instagram 35

clothes. White kids overall were posed more seductively and were dressed in more revealing

clothes than Black kids. This was mostly due to a difference between Black boys and White

boys, since girls were posed and dressed similarly with a very little effect of race. White boys

posed more seductively and exposed more skin in their photos while Black boys were more

likely to use dominant poses. This shows that teenagers are emulating the gender differences

seen in traditional media in their profile photos, so they are both picking up on gendered and

racialized messages in the media but also feel that they are meant to follow these standards.

The Current Study

The current study looks at gender presentation on Instagram while looking specifically at

the feedback that users got. Number of comments and number of likes are the most basic

measures of feedback and were measured in most of the previous studies. I focused on

comments, because comments give more information about what and why users think about

posts. Some previous studies included positive, negative, and neutral comments, which this study

also looks at. Comments that specifically reference gender presentation didn’t seem to be in any

previous studies, which is partially why I included that variable in this study. Whether or not

comments were about gender show if commenters were aware of the gender presentation when

giving positive or negative feedback. I chose Instagram because it is currently very popular

worldwide and focuses mainly on photo sharing. Other social medias may include photo sharing,

but that is not the main focus of the site. Instagram also encourages building a following outside

of family and friends an individual knows in “real life,” so the way a person presents themselves

might be completely different on Instagram than in person. It also is more similar to traditional

media, like advertisements, because a user is trying to sell themselves as a personality and may

even advertise products in their posts. Additionally, Instagram creates the parasocial
Gender and Instagram 36

relationships of traditional celebrities while giving a stronger illusion of closeness and personal

connection due to users addressing their audience as though they are friends. This is important,

because social media appears more authentic than traditional media so many don’t understand

how posed and calculated posts may be. It is also important, because social media is become

more intertwined with daily life and the images on social media are being imitated by young

people in real life, affecting the ways that entire generations conceptualize and present their

identities.

In my study, I also looked at gender presentation and gender roles on Instagram,

specifically looking at the popular pages of different hashtags. I looked at both more masculine-

leaning and more feminine-leaning hashtags as well as fairly gender-neutral ones all within

similar online communities. I expected that more masculine gender roles would be seen in posts

featuring men and more feminine gender roles would be seen in posts featuring women. I looked

at feedback on gendered posing and gender roles. I also looked at the difference between gender

roles and gender presentation on sponsored posts, unsponsored posts, and posts from business

accounts.

Hypotheses:

 Posts that use gendered posing that aligns with their perceived gender will receive

more positive feedback.

 More exaggerated versions of feminine posing will receive negative feedback.

 People will be more likely to police the gender presentation and gender roles in

posts that featuring people of their same gender.

 People of color will receive more negative feedback than white people.

 Comments will be related to the gender roles displayed in the posts.


Gender and Instagram 37

 When a post is more related to selling a product, it will also include more

stereotypical gender roles and gender posing.

Methods

Posts were selected by looking at the popular page every week under 6 different hashtags.

The first 50 Instagram posts near the top of the popular page every week for 6 weeks were

included. Any posts that were videos or didn’t feature a person were excluded from the sample.

The hashtags were chosen by identifying more masculine-leaning, feminine-leaning, and neutral

hashtags in similar categories. These hashtags were #menswear and #bodybuilding for more

masculine-leaning hashtags, #ootd and #fitnessmodel for more feminine-leaning hashtags, and

#streetwearoutfit and #fitnessgoals for neutral hashtags. Overall, 214 Instagram posts were

included in the sample.

The posts were then analyzed for different variables. This included setting, makeup,

feminine touch, ritualization of subordination, licensed withdrawal, body display, kissing pout,

muscle presentation, faceless portrayal, sponsored, gender, and race/ethnicity (see table below).

Gender was reported as male, female, or ambiguous based on apparent presentation. Race was

reported according to researcher’s judgement by looking at skin color and facial features. I also

combined race and gender into one variable to further look at the intersection of race and gender

with subjects of the post being categorized as a white woman, white man, black woman, black

man, etc. To look at the effect of the gender of the subject on the response gender posing

receives, another variable was made for each type of gender posing that included the gender of

the subject, whether or not they displayed the pose, and whether it was exaggerated. For

example, setting coded individuals as a woman in the home, a man in the home, a woman
Gender and Instagram 38

indoors, a man indoors, a woman outdoors, or a man outdoors. The unclear or androgynous

category was eliminated due to a small amount of posts fitting into the category.

Table 1

Description of Variables

Variable Types of Description Example

Variables

Setting In Home (1) Photo is taken inside a house.

Indoors not in Photo is taken indoors and is

Home (2) not in a house.


Gender and Instagram 39

Outdoors (3) Photo setting is outside.

Makeup Natural Makeup Subject is wearing natural,

(1) neutral colored makeup.

Heavy Makeup Subject is wearing heavy,

(2) dark, or brightly colored

makeup.
Gender and Instagram 40

No (3) Subject isn’t wearing makeup.

Body Display Yes (1) Subject is wearing revealing

clothing that would be

socially acceptable to wear in

public, such as short shorts or

an exposed midriff.

No (2) This is not featured in photo.

Exaggerated (3) Subject is wearing revealing

clothing that wouldn’t

typically be worn in public,

such as underwear or a

bathing suit.
Gender and Instagram 41

Kissing Pout Yes (1) Subject in photo is making a

kissing pout face in a natural

or subtle way.

No (2) This is not featured in photo.

Exaggerated (3) Subject in photo is making a

kissing pout face in an

unnatural or exaggerated way.


Gender and Instagram 42

Muscle Yes (1) Subject is posing in a way that

Presentation intentionally accentuated their

muscles in a way that still

looks somewhat natural.

No (2) This is not featured in photo.

Exaggerated (3) Subject posed in a way that

intentionally accentuates their

muscles in an unnatural or

exaggerated way.

Sponsored Yes (1) Post is advertising a product

or company.

No (2) Post doesn’t feature any

advertisements.
Gender and Instagram 43

Company account Post is from a company

(3) account.

Faceless Yes (1) Subject of photo’s face or

Portrayal head is hidden or cropped out.

No (2) This is not featured in photo.

Licensed Yes (1) Subject of photo is closing

Withdrawal their eyes or looking away or

is showing loss of control of

emotion (example: expansive

smile, loud laughter, covering

face, covering mouth, biting

fingers)
Gender and Instagram 44

No (2) This is not featured in photo.

Ritualization Yes (1) Subject of photo is canting

of their head or body, standing

Subordination on one foot, crossing legs,

leaning on others for support,

sitting, or lying down

No (2) This is not featured in photo


Gender and Instagram 45

Feminine Yes (1) Subject of photo is touching

Touch clothes or themselves or their

hands are tracing an object

No (2) This is not featured in the

photo

Comment Positive (1) Comment intended to be

type positive

Negative (2) Comment intended to be

negative

Unclear (3)

No comment (4)

Commenter Male (1) Commenter appears to be

Gender male based on profile photo

Female (2) Commenter appears to be

female based on profile photo


Gender and Instagram 46

Unclear (3) Gender of commenter not

apparent

No comment (4)

Commenting Yes (1) Comment directly references

on Gender the gender presentation or

gender roles shown in photo

(usually

discussing/sexualizing the

body or the clothes someone

is wearing, emojis by

themselves are a no unless its

muscles)

No (2) Comment not related to

gender

No comment (3)

The top five comments underneath each post were also recorded, with 159 comments

recorded overall. Any comments that weren’t in English or that were clearly spam were excluded

from the sample. Whether the commenter was male or female or unknown, whether the comment

was positive or negative, and whether the comment directly commented on the gender

presentation and gender roles shown in the photo were recorded. Gender of the commenter was
Gender and Instagram 47

posted by looking at the apparent presentation of the commenter in the profile photo and was

reported as male, female, or ambiguous.

Results.

The demographics of the posts are mentioned first with table 2 showing the gender of the

subjects in the photos and table 3 showing the race and ethnicity of subjects in the photos. Next

is the engagement with table 4 showing the number of comments and table 5 showing the

number of likes. Table 6 shows whether posts are sponsored or non-sponsored or from a

company account. Next, the content of the posts, specifically the gendered content, with table 7

showing the setting of the posts and table 8 showing the gendered posing. Next is the content of

the comments with tables 9-13 showing the types of comments. Lastly, table 14 shows the

gender of the commenter.

Table 2

Gender of Subjects in Photos

Gender of subject Number of posts Percent of posts

Male 88 41.1%

Female 118 55.1%

Table 3

Race and Ethnicity of Subjects in Photos

Race/Ethnicity of subject Number of posts Percent of posts

Black 23 10.7%

White 137 64.0%

Asian 23 10.7%
Gender and Instagram 48

Latinx 7 3.3%

Ambiguous/Unclear 24 11.2%

Table 4

Number of Comments on Posts

Number of comments Number of posts Percent of Posts

0-10 comments 57 26.6%

11-20 comments 24 11.2%

21-30 comments 28 13.1%

31-40 comments 12 5.6%

41-50 comments 15 5.6%

51-60 comments 7 3.3%

61-70 comments 5 3.3%

71-80 comments 7 3.3%

81-90 comments 6 2.8%

91-100 comments 7 3.3%

Over 100 comments 46 21.5%

Table 5

Number of Likes on Posts

Number of likes Number of posts Percent of posts


Gender and Instagram 49

499 or less likes 99 46.3%

500-999 likes 31 14.5%

1000-1999 likes 32 15.0%

2000-3999 likes 19 8.9%

4000 or more likes 18 8.4%

Table 6

Sponsored and Non-Sponsored Posts

Type of Post Number of posts Percent of posts

Sponsored Post 35 16.4%

Company Account 21 9.8%

Non-sponsored posts 158 73.8%

Table 7

Setting of Posts

Type of Setting Number of Posts Percent of Posts

In the Home 38 17.8%

Indoors 75 35.0%

Outdoors 101 47.2%

Table 8

Gendered Posing

Gender Posing Number of Posts Percent of Posts


Gender and Instagram 50

Natural Makeup 64 29.9%

Heavy Makeup 25 11.7%

No Makeup 125 58.4%

Body Display Present 61 28.5%

Exaggerated Body Display 31 14.5%

No Body Display 122 57.0%

Kissing Pout Present 21 9.8%

Exaggerated Kissing Pout 2 0.9%

No Kissing Pout 191 89.3%

Muscle Presentation Present 48 22.4%

Exaggerated Muscle 24 11.2%

Presentation

No Muscle Presentation 142 66.4%

Faceless Portrayal Present 27 12.6%

No Faceless Portrayal 187 87.4%

Licensed Withdrawal Present 103 48.1%

Licensed Withdrawal Not 111 51.9%

Present

Ritualization of 38 17.8%

Subordination Present

Ritualization of 176 82.2%

Subordination Not Present

Feminine Touch Present 40 18.7%


Gender and Instagram 51

Feminine Touch Not Present 174 81.3%

Table 9

Types of First Comments

Type of Comment Number of Comments Percent of Comments

Positive 117 (82.7%)

Negative 1 (.5%)

Unclear 13 (6.1%)

About Gender 73 (34.1%)

Not About Gender 117 (54.7%)

Table 10

Types of Second Comments

Type of Comments Number of Comments Percent of Comments

Positive 162 (75.7%)

Negative 1 (.5%)

Unclear 16 (7.5%)

About Gender 57 (26.6%)

Not About Gender 122 (57.0%)

Table 11

Types of Third Comments

Type of Comments Number of Comments Percent of Comments


Gender and Instagram 52

Positive 158 (73.8%)

Negative 2 (.9%)

Unclear 12 (5.6%)

About Gender 63 (29.4%)

Not About Gender 109 (50.9%)

Table 12

Types of Fourth Comments

Type of Comments Number of Comments Percent of Comments

Positive 158 (73.8%)

Negative 2 (.9%)

Unclear 6 (2.8%)

About Gender 71 (33.2%)

Not About Gender 95 (44.4%)

Table 13

Types of Fifth Comments

Type of Comments Number of Comments Percent of Comments

Positive 150 (70.1%)

Negative 2 (.9%)

Unclear 7 (3.3%)

About Gender 62 (29.0%)

Not About Gender 97 (45.3%)


Gender and Instagram 53

Table 14

Gender of the Commenter

Commenter Gender Number of Commenters Percent of Commenters

Male 374 43.14%

Female 336 38.75%

Unclear 157 18.11%

To look at general patterns in the data, I looked at the relationship between gender and

number of likes and between race and number of likes using an independent sample t test for

each of them. There weren’t any significant results.

To answer my research questions about Instagram, gender, and feedback I looked at

specific relationships within gendered posing, sponsored posts, and comments. Within gendered

posing I looked at every individual gendered pose-feminine touch, makeup, muscle presentation,

body display, kissing pout, faceless portrayal, licensed withdrawal, and ritualization of

subordination. I looked at the relationship between the gendered posing and the gender of the

subject in the photo as well as between the setting of the photo and the gender of the subject.

This was done through chi squares for each individual gendered pose and for setting. Gender had

a significant relationship with makeup (X2 (2, N = 206) = 116.862, p = .000.), kissing pout (X2

(2, N = 206) = 7.910, p = .019.), licensed withdrawal (X2 (1, N = 206) = 10.895, p = .001.),

ritualization of subordination (X2 (1, N = 206) = 4.982, p = .026.), feminine touch (X2 (1, N =

206) = 19.734, p = .000.), and setting (X2 (2, N = 206) = 20.169, p = .000.). Women wore

makeup more than men (N=0), and they wore natural makeup (N=64) more than heavy makeup
Gender and Instagram 54

(N=25). Women displayed kissing pout (N=16) more than men (N=3). Women were more likely

to not display kissing pout (N=100) than display kissing pout (N=16), and they were more likely

to display kissing pout normally than in an exaggerated way (N=2). Men were more likely to

display licensed withdrawal (N=54) than women (N=45). Women displayed ritualization of

subordination (N=26) more than men (N=9). Women displayed feminine touch (N=34) more

than men (N=4), but women did not display feminine touch (N=84) more than they did. Women

were outdoors more than in any other setting (N=44), but men were more likely to be outdoors

than women (N=51). Next, I looked at the relationship between gendered posing-both whether or

not it was present and if it was exaggerated- and likes by running a one-way ANOVA, and I

looked at the relationship between gender posing and number of comments also by running a

one-way ANOVA (what kind). There weren’t any significant results.

To further examine the relationship between gendered posing, number of likes, and

number of comments I looked at how that varied depending on the gender of the poster and the

race or ethnicity of the poster. First, I looked at the relationship between race, gender, and gender

posing by running a chi square with race and gender as one variable and each kind of gender

posing as the other. Makeup was the only gendered pose with a significant relationship (X2 (18,

N = 8) = 141.308, p = .000.). White women wore natural makeup more than any group (N=43),

and they wore heavy makeup more than any group (N=17). Latinx women were the only group

that wore heavy makeup (N=4) more than natural (N=1) or no makeup (N=1). Then with the

same race and gender variable I ran a chi square with the setting of the photo. This relationship

was significant (X2 (18, N = 8) = 39.730, p = .002.). White women were in the home (N=21) and

indoors (N=31) more than other groups, and white men were outdoors more than other groups
Gender and Instagram 55

(N=35). Black men (N=7), black women (N=4), Asian men (N=8), and Asian women (N=6)

were more likely to be outdoors than in other settings.

Then, I made a variable for both gendered posing and gender of the subject to look at the

relationship between the gender of the subject, gendered posing (including setting), and number

of likes with a one-way ANOVA. There weren’t any significant results. I used the same variable

to run a one-way ANOVA with number of comments. This relationship was significant (F (127,

86) =1.622, p = .009.). Men outdoors had the most posts in the 100+ comment category (N=15),

but they also had the most posts in the less than 10 comment category (N=14). Men inside

(N=12), women outside (N=11), and women inside (N=11) had more posts in the less than 10

comment category than in other categories. Women without makeup had the most posts in the

100+ comment category (N=18). Women with exaggerated makeup also had the most posts in

the less than 10 comment category (N=14). Men that didn’t use body display had the most posts

in the 100+ comment category (N=15), but they also had the most posts in the less than 10

comment category (N=17). Women that didn’t display muscle presentation had the most posts in

the 100+ comment category (N=18). Men without muscle presentation had the most posts in the

less than 10 comment category (N=18). Men that didn’t display kissing pout had the most posts

in the 100+ comment category (N=21), and they also had the most posts in the less than ten

comment category (N=26). Women who didn’t use faceless portrayal had the most posts in the

100+ comment category (N=21). Men who didn’t use faceless portrayal had the most posts in the

less than 10 comment category (N=23). Men who used licensed withdrawal and women who

used licensed withdrawal both had the most posts in the 100+ comment category (N=12). Men

who used licensed withdrawal had the most posts in the less than 10 comment category (N=19).

Woman who didn’t display ritualization of subordination had the most posts in the 100+
Gender and Instagram 56

comment category (N=20). Men who didn’t display ritualization of subordination had the most

posts in the less than 10 comment category (N=25). Men who didn’t display feminine touch had

the most posts in the 100+ comment category (N=20), but they also had the most posts in the less

than 10 comment category (N=25). I then made a variable for gendered posing and the race or

ethnicity of the subject. I used this variable to run a one-way ANOVA with number of likes and

then another one-way ANOVA with number of comments. There weren’t any significant results.

I then looked at how sponsored posts changed presentation and feedback. I first looked at

the relationship between whether or not a post was sponsored and gendered posing with a chi

square. Only the relationship with muscle presentation was significant (X2 (4, N = 214) = 9.703,

p = .046.). Not displaying muscle presentation was the most common in the sponsored (N=22),

unsponsored (N=100), and company account (N=20) categories. I also ran a chi square with

sponsored posts and setting of the photo. There weren’t any significant results. Next, I ran a chi

square with race or ethnicity of the subject and sponsored posts to see if there was a difference in

who was featured in sponsored posts. There weren’t any significant results. Then I looked at the

relationship between sponsored posts, gendered posing, and race with a chi square. There

weren’t any significant results. I also looked at the relationship between sponsored posts, race,

and setting with another chi square. There weren’t any significant results. Then, I ran a chi

square with gender posing and the gender of the subject and whether or not a post was

sponsored. There weren’t any significant results. I also ran a chi square to look at the relationship

between setting, the gender of the subject, and whether or not a post was sponsored. This

relationship was significant (X2 (10, N = 206) = 25.847, p = .004.). Women inside (N=11) and

men outside (N=5) were most common in the sponsored category. Women in the home (N=28)
Gender and Instagram 57

and men outside (N=43) were most common in the not sponsored category. Women outside

(N=8) and men inside (N=6) were most common in the company account category.

To further explore the feedback these posts got, I looked at the comments, more

specifically the types of comments. Each analysis was done five times, once for each of the first

five comments. First, I looked at the relationship between the gender of the subject and whether a

comment was negative or positive with a chi square. This was only significant with the second of

the first five comments on every post (X2 (2, N = 172) = 6.775, p = .034.). Women were more

likely to receive positive comments (N=95), and while men still receive more positive comments

than other comments (N=60), they received more vague or unclear comments (N=11) than

women (N=5). I also looked at the relationship between the gender of the commenter and

whether the comment was negative or positive with another chi square. There weren’t any

significant results. I then looked at the relationship between the race or ethnicity of the subject

and whether comments were positive or negative with a chi square. There weren’t any significant

results. I also ran a chi square to see the relationship between comments specifically about

gender and race or ethnicity of the subject. There weren’t any significant results.

I then looked at the relationship between the gender of the commenter, the gender of the

subject, and whether comments were positive or negative by running a chi square. This was

significant for the third comment only (X2 (8, N = 140) = 16.861, p = .032.). Women who were

commenting on women made the most positive comments (N=45), and men who were

commenting on women made the most unclear comments (N=4). I also ran a chi square to look

at the relationship between the gender of the commenter, the gender of the subject, and whether

the comment was about gender. This was significant for the first (X2 (6, N = 146) = 20.013, p =

.003.) and second (X2 (4, N = 140) = 13.829, p = .008.) comments. Women commenting on
Gender and Instagram 58

women made more comments on gender (N=29) for comments 1 and 2. I then ran a chi square to

look at the relationship between gendered posing, gender of the subject, and whether comments

were negative or positive, and I also ran a chi square with setting of the photo, gender of the

subject, and whether comments were positive or negative. This was significant when looking at

the first of the first five comments of every post and body display (X2 (10, N = 184) = 24.971, p

= .005.). Women outside received the most positive comments (N=36), and men inside received

the most unclear comments (N=5). This was significant with the muscle presentation gendered

pose and the first of the five comments (X2 (10, N = 184) = 20.006, p = .011). Women who

didn’t display muscle presentation received the most positive comments (N=70), and men who

didn’t display muscle presentation received the most unclear comments (N=5). This was also

significant with the second comment and the kissing pout gendered pose (X2 (8, N = 172) =

17.241, p = .028.). Women who didn’t display kissing pout received the most positive comments

(N=78), and men who didn’t display kissing pout received the most unclear comments (N=9).

The third comment was significant with the body display (X2 (10, N = 165) = 34.563, p = .000.)

and muscle presentation poses (X2 (10, N = 165) = 37.665, p = .000.). Women who didn’t use

body display received the most positive comments (N=45). Women who didn’t use muscle

presentation received the most positive comments (N=63). The fourth comment and the faceless

portrayal gendered pose (X2 (6, N = 160) = 21.706, p = .001.).

I also looked at the relationship between the gendered posing, the gender of the subject,

and whether the comment was about gender by running a chi square. The first comment and

licensed withdrawal had a barely significant relationship (X2 (6, N = 200) = 12.459, p = .052).

Women who didn’t display licensed withdrawal received the most comments about gender

(N=30), and men who did display licensed withdrawal received the most comments not about
Gender and Instagram 59

gender (N=37). The first comment and the setting were barely significant (X2 (10, N = 200) =

18.138, p = .053.). Women who were indoors received the most comments related to gender

(N=17), and women outside received the most comments not related to gender (N=23). The

second comment and ritualization of subordination were significant (X2 (6, N = 197) = 15.819, p

= .015.). Women who didn’t display ritualization of subordination received the most comments

on gender (N=36), and men who didn’t display ritualization of subordination received the most

comments not about gender (N=50). Lastly, I ran a chi square to look at the relationship between

the gender of the subject, the setting of the photo, and whether or not the comments were about

gender. This was only significant with the first comment of each post. Women inside received

the most gendered comments (N=17), and women outside received the most comments not about

gender (N=23).

Discussion

Overall, who conformed to gender roles the most were the most interesting findings.

Women displayed feminine gender roles more than men displayed feminine gender roles, and

white people conformed to traditional gender roles the most. Additionally, non-sponsored posts

conformed more to ideas of traditional masculine and feminine spheres than sponsored posts and

company accounts. Another interesting and unexpected finding was that the number of

comments seemed to be a better measure of engagement than number of likes. Number of likes

never had any significant results, but number of comments showed many interesting patterns.

In order to see how gender is presented on Instagram, I first looked at the gendered

posing that has been seen in previous studies. I found that settings of the photos and whether or

not makeup was featured has a significant relationship with gender by itself. Men were outdoors

more than women, following traditional gender roles where women are confined to the home
Gender and Instagram 60

while men are not. Women wore makeup more than men, and they were more likely to wear

natural makeup than heavy makeup. Heavy makeup can be seen as an exaggeration of makeup as

gender presentation, so this could mean that women are hesitant to exaggerate feminine gender

roles. This is similar to what was found in Butkowski et al (2020) where, in comparison to

traditional media, gender cues were implemented in subtle and de-emphasized ways on social

media to give the appearance of effortlessly performing gender.

Gender also had significant relationships with kissing pout, licensed withdrawal,

ritualization of subordination, and feminine touch. Women displayed kissing pout, feminine

touch, and ritualization of subordination more than men did. This shows that overall women are

more likely to display gender posing associated with their gender. Men were more likely to

display licensed withdrawal than women, the only finding where men display feminine gender

roles more than women do. I noticed that many of these posts were from photos meant to be

showing off style and fashion sense. This could mean that certain poses tend to be popular in

different online communities or that strict gender role adherence isn’t as highly valued in some

online communities.

General gender posing in a photo, including setting, was significant with the number of

comments. This shows that using gendered posing encourages interaction with Instagram posts.

This could also encourage users to use gendered posing, because more interaction with Instagram

posts leads to the algorithm showing the posts to more users, which could lead to more followers

as well as likes and comments in the future. This could help users by also getting them more

sponsorships or get paid more in sponsorships.

There were some particularly interesting and unexpected findings with gender, gendered

posing, and number of comments. Women who didn’t wear makeup received more comments
Gender and Instagram 61

than women who wore natural makeup, while women who wear heavy or exaggerated makeup

received the least comments. This shows that women who show exaggerated feminine roles

receive the least engagement, which might support the idea that exaggerated feminine roles tend

to receive negative feedback. Women not wearing makeup received more engagement, which is

interesting since that doesn’t conform to feminine gender roles. This could be due to general

controversy online over women who wear a lot of makeup. They are sometimes viewed as fake

or lying, which can be detrimental to an online relationship with an audience that relies on

feelings of authenticity and closeness. Women who didn’t display muscle presentation received

more comments while men who didn’t display muscle presentation received the least comments.

This could mean that in general, people who follow traditional gender roles for their gender

receive the most comments, usually positive reinforcement since most of the comments are

positive, while people who don’t receive very little engagement. It’s also interesting to point out

that it is slightly more acceptable for women to behave more masculinely, partially due to

feminism and partially due to the general valuing of masculinity. Women receiving more

positive feedback for following gender roles and men receiving the least feedback for not

following gender roles could be due to this. Women who didn’t use faceless portrayal received

the most comments, and men who didn’t use faceless portrayal received the least comments. This

is interesting, because faceless portrayal is considered feminine making this finding may support

the opposite of the previous one. Women who didn’t display ritualization of subordination

received the most comments while men who didn’t display ritualization of subordination

received the least comments. This finding is similar to the relationship between faceless

portrayal, gender, and number of comments. It’s unclear why number of comments have the

relationship with faceless portrayal and ritualization of subordination, but I think it may be
Gender and Instagram 62

related to objectification. Both faceless portrayal and ritualization of subordination seem to be

related to objectification that women tend to experience. While muscle presentation may also be

objectifying, it isn’t usually in the submissive way that is typical of feminine gender roles.

The relationship between number of comments supported the hypothesis that gender

posing aligning with a subject’s gender receives more positive feedback. I previously

hypothesized that exaggerated feminine poses would receive more negative comments, but no

significant findings were found to support this. This could be due to the lack of negative

comments overall, making it difficult to find any significant relationship with negative

comments.

There were no significant relationships with number of likes. This might mean that

number of likes isn’t an accurate way of measuring engagement or feedback. This is interesting

given that most studies involving social media feedback use number of likes to measure

feedback, and Instagram itself uses number of likes in its algorithm.

There were a few findings when looking at the intersection of race and gender that were

similar to relationships between gender and gender posing in the posts. Race and gender had a

significant relationship with the setting of photos and with whether or not makeup was featured

in a post. White women wore both natural and heavy makeup more than any group. This might

mean that white women are more likely to strictly follow gender roles by wearing makeup.

White women might feel more comfortable displaying exaggerated gender roles, like heavy

makeup, due to having their gender presentation less heavily policed by others than women of

color. Additionally, it is easier to find more natural makeup when the selection is limited, so

women of color might have a more difficult time finding makeup that would appear exaggerated

and work with their skin tone. No men were displayed wearing makeup regardless of race or
Gender and Instagram 63

ethnicity, showing men strictly following traditional gender roles. The setting again showed

white women following traditional gender roles more strictly with white women being in the

home and indoors more than other groups. White men were outdoors more than other groups,

showing white men more strictly following traditional gender roles. It seems that white people

might follow traditional gender roles more than people of color. This might be because ideas

about traditional gender roles in the United States were originally defined and determined by

white people.

Unlike in the intersection of race and gender, there were no tests where race alone was

significant. This shows that while previous findings about gender presentation often ignored

people of color, the ways in which gender is presented on Instagram isn’t only seen in white

people. This is also especially interesting in the context of these all being the top posts. This

could mean that Instagram promotes the same type of content regardless of race. The top posts

have been found to emulate each other in Baker and Walsh (2018), which found that the styles

and forms of top images under a certain hashtag are usually similar giving communities on

Instagram a cohesive look. They also suggested that this was done purposefully by users to gain

recognition and approval. In the context of this previous finding, this might just mean that across

demographics the same content is being reproduced, creating a homogenous culture and

homogenous “look” of Instagram as a whole. This might be something that happened in recent

years since the previous study where Instagram itself is becoming a larger cohesive community.

Whether this is true among less popular posts is hard to determine when only looking at the top

posts rather than a sampling that is more indicative of the user base as a whole.

Race and gender didn’t have as much of a relationship when it came to feedback. I had

previously hypothesized that people of color would receive more negative comments, but this
Gender and Instagram 64

finding was not significant. This could possibly be due to a general lack of negative comments

overall. The lack of negative comments, however, could be due to users filtering their comments

and removing any frequent negative comments they receive.

This makes looking at sponsored posts and business accounts even more important,

because they may be a driving factor in the types of photos users’ post. The number of comments

was also seen to have a significant relationship with whether or not a post was sponsored. This

further feeds the idea that number of comments, and engagement in general, could be

incentivizing users to create certain content in order to receive more sponsored posts. This could

also mean that companies and brands that sponsor Instagram users could be asking them to

present themselves in a certain way, because they know it will get more comments and more

engagement which leads to more people seeing the sponsored post via the algorithm.

I had previously hypothesized that gender posing would be different in sponsored posts

than in regular non-sponsored posts. I found that whether or not a post was sponsored had a

significant relationship with the gender of the poster and the setting of the post. In sponsored

posts, women were shown more inside but not in the home while men were shown outside more,

and in posts on company accounts, women were more likely to be outside and men were more

likely to be inside. Non-sponsored posts were more likely to show women in the home and men

outside, the settings that conform to traditional gender roles the most. Settings that reflect

traditional masculine and feminine domains might be more comfortable for people creating their

own non-sponsored post, while a brand might ask for specific settings and poses they believe

might boost engagement or be more persuasive in a brand or company’s marketing.

The comments, specifically whether comments were positive or negative and whether

comments were about gender, were the most confusing and inconsistent findings. I had
Gender and Instagram 65

hypothesized that commenters were more likely to comment on the gender presentation of users

of the same gender. There were some significant findings for commenter gender and the gender

of the poster. For the first and second comments of every post, comments on gender and the

gender of the poster were significant. These inconsistent findings could partially be due to any

missing comments being the last few, so the sample was smaller for these comments.

I had previously hypothesized that there would be comments about the gender roles

portrayed in the posts. This was measured through the commenting on gender variable. The first

five comments of each post were looked at to measure this. 34.1% of the first comments were

about gender, 26.6% of the second comments were about gender, 29.4% of the third comments

were about gender, 33.2% of the fourth comments were about gender, and 29.0% of the fifth

comments were about gender. Part of the reason there weren’t many comments about gender

could be that many of the comments didn’t include any words and were instead entirely made up

of emojis. This made it difficult to figure out if comments were negative or positive as well as if

they were about gender or not. If I were to look at comments again, I would leave out comments

without words in the sample.

I think that overall, the methods I used were effective, but they only gave answers about

the top posts. This means that the patterns I found mostly have to do with what is popular and

what Instagram promotes instead of what the general userbase is posting. However, this does tell

us what most people are seeing on Instagram and what they are interacting with.

A future study should combine all comments instead of making each of the five

comments a different variable, though there is enough data about comments to do a study only

on Instagram comments. Additionally, future studies should look at a sample that is more

representative of all Instagram posts, instead of just the most popular posts. This would give
Gender and Instagram 66

more information about average posts and help determine how much of the patterns have to do

with what is promoted by the algorithm versus what is actually common among Instagram users.
Gender and Instagram 67

References

Baker, S. A., & Walsh, M. J. (2018). ‘Good Morning Fitfam’: Top posts, hashtags and gender

display on Instagram. New Media & Society, 20(12), 4553–4570.

https://doi.org/10.1177/1461444818777514

Bazzini, D. G., Pepper, A., Swofford, R., & Cochran, K. (2015). How healthy are health

magazines? A comparative content analysis of cover captions and images of Women’s and

Men’s Health magazine. Sex Roles: A Journal of Research, 72(5–6), 198–210.

https://doi.org/10.1007/s11199-015-0456-2

Bradley, K. (2013). (Re)presentations of (hetero)sexualized gender in Two and a Half Men : a

content analysis. Journal of Gender Studies, 22(2), 221–226.

https://doi.org/10.1080/09589236.2012.752348

Bresnahan, M. J., Inoue, Y., Liu, W. Y., & Nishida, T. (2001). Changing gender roles in prime-

time commercials in Malaysia, Japan, Taiwan, and the United States. Sex Roles: A Journal

of Research, 45(1–2), 117–131. https://doi.org/10.1023/A:1013068519583

Butkowski, C. P., Dixon, T. L., Weeks, K. R., & Smith, M. A. (2020). Quantifying the feminine

self(ie): Gender display and social media feedback in young women’s Instagram selfies.

New Media & Society, 22(5), 817–837. https://doi.org/10.1177/1461444819871669

Dhillon, Sunny. “How Instagram Is Eating The World.” Forbes, Forbes Magazine, 25 June 2018,

www.forbes.com/sites/valleyvoices/2018/06/25/how-instagram-is-eating-the-

world/?sh=1a682b703145.

Döring, N., Reif, A., & Poeschl, S. (2016). How gender-stereotypical are selfies? A content

analysis and comparison with magazine adverts. Computers in Human Behavior, 55(Part

B), 955–962. https://doi.org/10.1016/j.chb.2015.10.001


Gender and Instagram 68

Elizabeth Behm-Morawitz (2017) Examining the intersection of race and gender in video game

advertising, Journal of Marketing Communications, 23:3, 220-

239, DOI: 10.1080/13527266.2014.914562

Everett, B.G., Steele, S.M., Matthews, A.K. et al. Gender, Race, and Minority Stress Among

Sexual Minority Women: An Intersectional Approach. Arch Sex Behav 48, 1505–1517

(2019). https://doi.org/10.1007/s10508-019-1421-x

“Facebook vs Instagram: Where Should You Spend Your Marketing Monnies in 2021.” Business

Insider India, Business Insider India, 24 Mar. 2021,

www.businessinsider.in/advertising/ad-tech/article/facebook-vs-instagram-where-should-

you-spend-your-marketing-monnies-in-2021/articleshow/81327753.cms.

Furnham, A., & Bitar, N. (1993). The stereotyped portrayal of men and women in British

television advertisements. Sex Roles: A Journal of Research, 29(3–4), 297–310.

https://doi.org/10.1007/BF00289940

Furnham, A., & Skae, E. (1997). Changes in the stereotypical portrayal of men and women in

British television advertisements. European Psychologist, 2(1), 44–51.

https://doi.org/10.1027/1016-9040.2.1.44

Gaa, J. P., Liberman, D., & Edwards, T. A. (1979). Journal of Clinical Psychology, 35(3), 592–

598. https://doi.org/10.1002/1097-4679(197907)35:3<592::AID-

JCLP2270350322>3.0.CO;2-W

Goffman, E. (1977). The Arrangement between the Sexes. Theory & Society, 4(3), 301.

https://doi.org/10.1007/BF00206983
Gender and Instagram 69

Hancock, B.H. and Garner, R. (2015), Erving Goffman: Theorizing the Self in Advanced

Consumer Capitalism. J Theory Soc Behav, 45: 163-187.

https://doi.org/10.1111/jtsb.12062

Hawke, L. A. (2007). Gender roles within American marriage: Are they really

changing?. Essai, 5(1), 23.

Hershey, M. R. (1978). Social Science Quarterly (University of Texas Press), 58(4), 583–596.

Kapidzic, S., & Herring, S. C. (2015). Race, gender, and self-presentation in teen profile

photographs. New Media & Society, 17(6), 958–976.

https://doi.org/10.1177/1461444813520301

Park, H., Rodgers, S., McElroy, J. A., & Everett, K. (2018). Sexual and gender minority’s social

media user characteristics: Examining preferred health information. Health Marketing

Quarterly, 35(1), 1–17. https://doi.org/10.1080/07359683.2017.1310553

Plous, S., & Neptune, D. (1997). Racial and Gender Biases in Magazine Advertising. Psychology

of Women Quarterly, 21(4), 627. https://doi.org/10.1111/j.1471-6402.1997.tb00135.x

Puglisi, J. T. (1980). Journal of Personality Assessment, 44(3), 272.

https://doi.org/10.1207/s15327752jpa4403_9

Rubie-Davies, C. M., Liu, S., & Lee, K. C. K. (2013). Watching each other: Portrayals of gender

and ethnicity in television advertisements. The Journal of social psychology, 153(2), 175-

195.

Smith, L. R., & Sanderson, J. (2015). I’m Going to Instagram It! An Analysis of Athlete Self-

Presentation on Instagram. Journal of Broadcasting & Electronic Media, 59(2), 342–358.

https://doi.org/10.1080/08838151.2015.1029125
Gender and Instagram 70

Song, J., Han, K., Lee, D., & Kim, S.-W. (2018). “Is a picture really worth a thousand words?”: A

case study on classifying user attributes on Instagram. PLoS ONE, 13(10), 1–22.

https://doi.org/10.1371/journal.pone.0204938

Suciu, Peter. “Is Instagram The Social Media Service For Business In 2020?” Forbes, Forbes

Magazine, 24 Dec. 2019, www.forbes.com/sites/petersuciu/2019/12/26/is-instagram-the-

social-media-service-for-business-in-2020/?sh=4d48450f3bdf.

Vrugt, A., & Van Eechoud, M. (2002). Smiling and self-presentation of men and women for job

photographs. European Journal of Social Psychology, 32(3), 419–431.

https://doi.org/10.1002/ejsp.99
Gender and Instagram 71

Appendix

I ran a chi square with the race and gender variable and gendered posing. Body display

wasn’t significant, X2 (18, N = 8) = 20.553, p = .303. Kissing pout wasn’t significant, X2 (18, N

= 8) = 22.553, p = .208. Muscle presentation wasn’t significant, X2 (18, N = 8) = 14.786, p =

.677.

I ran an independent sample t test to see the relationship between gender and likes which

turned out to not be significant, t(191) = 1.302, p = .194.

I ran an ANOVA to see the relationship between race and likes, which also had no

significant results, F (4, 194) = 2.262, p = .064. I also ran an ANVOA to see the relationship

between gender and likes, which was also not significant, F (1, 191) = .780, p = .378.

I ran ANOVAS to see the relationship between likes and gender posing, including

setting, (F (120, 78) = 1.080, p = .360.) and to see the relationship between number of comments

and gender posing (F (125, 79) =1.337, p = .082.). Neither ANOVA found any significant

results.

I ran a one-way ANOVA to see the relationship between the gender poster and gender

posing variables and likes. There were no significant relationships, F (120, 71) =.797, p = .863.

I ran chi squares to see how whether or not a post was sponsored affected makeup (X2 (4,

N = 214) = 2.436, p = .656.), body display (X2 (4, N = 214) = 6.677, p = .154.), kissing pout (X2

(4, N = 214) = 3.309, p = .507.), faceless portrayal (X2 (2, N = 214) = 2.365, p = .306.), licensed

withdrawal (X2 (2, N = 214) = 2.523, p = .283.), ritualization of subordination (X2 (2, N = 214)

= .193, p = .908.), feminine touch (X2 (2, N = 214) = 3.285, p = .194.), and setting (X2 (4, N =

214) = 3.578, p = .466.). None of these relationships were significant.


Gender and Instagram 72

I then ran chi squares to see how whether or not a post was sponsored affected feminine

touch and gender poster (X2 (6, N = 206) = 6.501, p = .369.), ritualization of subordination and

gender poster (X2 (6, N = 206) = 3.048, p = .803.), licensed withdrawal and gender poster (X2

(6, N = 206) = 7.573, p = .271.), faceless portrayal and gender poster (X2 (6, N = 206) = 5.816, p

= .444.), kissing pout and gender poster (X2 (6, N = 206) = 6.273, p = .444.), body display and

gender poster (X2 (6, N = 206) = 11.889, p = .293.), makeup and gender poster (X2 (6, N = 206)

= 4.589, p = .597.), and muscle presentation and gender poster (X2 (6, N = 206) = 13.452, p =

.199.). None of these relationships were significant.

I also ran a chi square to see the relationship between whether or not a post was

sponsored and the race or ethnicity of the poster (X2 (8, N = 214) = 9.633, p = .292.), which

wasn’t significant. I then ran chi squares to see how whether or not a post was sponsored affected

muscle presentation and race (X2 (28, N = 214) = .055, p = .426.), makeup and race (X2 (28, N =

214) = 32.332, p = .261.), body display and race (X2 (28, N = 214) = 28.276, p = .558.), kissing

pout and race (X2 (22, N = 214) = 16.112, p = .810.), faceless portrayal and race (X2 (16, N =

214) = 16.405, p = .425.), licensed withdrawal and race (X2 (18, N = 214) = 23.868, p = .159.),

ritualization of subordination and race (X2 (18, N = 214) = 24.092, p = .152.), feminine touch

and race (X2 (18, N = 214) = 18.584, p = .418.), and setting and race (X2 (28, N = 214) =

25.981, p = .574.). None of the relationships were significant.

I ran chi squares to see the relationship between gender and gender posing variables.

Body display (X2 (2, N = 206) = 4.042, p = .132.), muscle presentation (X2 (2, N = 206) = .632,

p = .729.), and faceless portrayal (X2 (1, N = 206) = .302, p = .582.) didn’t have significant

relationships with gender of the poster.


Gender and Instagram 73

I then ran chi squares to see the relationship between whether the first comments were

positive or negative and feminine touch and gender poster (X2 (6, N = 184) = 4.704, p = .582.),

ritualization of subordination and gender poster (X2 (6, N = 184) = 5.484, p = .483.), licensed

withdrawal and gender poster (X2 (6, N = 184) = 5.783, p = .448.), faceless portrayal and gender

poster (X2 (6, N = 184) = 7.050, p = .316.), kissing pout and gender poster (X2 (8, N = 184) =

7.000, p = .537.), makeup and gender poster (X2 (6, N = 184) = 4.215, p = .648.), setting and

gender poster (X2 (10, N = 184) = 17.347, p = .067.). None of these relationships were

significant.

I ran chi squares to see the relationship between whether the second comments were

positive or negative and feminine touch and gender (X2 (6, N = 184) = 8.320, p = .216.),

ritualization of subordination and gender (X2 (6, N = 172) = 8.001, p = .238.), licensed

withdrawal and gender (X2 (6, N = 172) = 8.997, p = .174.), faceless portrayal and gender (X2

(6, N = 172) = 9.031, p = .172), body display and gender (X2 (10, N = 172) = 16.132, p = .096.),

makeup and gender (X2 (6, N = 172) = 7.381, p = .287.), muscle presentation and gender (X2

(10, N = 172) = 17.646, p = .061.), and setting and gender (X2 (10, N = 172) = 16.961, p =

.075.). None of these relationships were significant.

I also ran chi squares to see the relationship between whether the third comments were

positive or negative and feminine touch and gender poster (X2 (6, N = 165) = 4.137, p = .658.),

ritualization of subordination and gender poster (X2 (6, N = 165) = 4.881, p = .559.), licensed

withdrawal and gender poster (X2 (6, N = 165) = 5.403, p = .493.), faceless portrayal and gender

poster (X2 (6, N = 165) = 9.379, p = .153.), kissing pout and gender poster (X2 (8, N = 165) =

6.720, p = .567.), makeup and gender poster (X2 (6, N = 165) = 3.802, p = .703.), and setting and
Gender and Instagram 74

gender poster (X2 (10, N = 165) = 17.672, p = .061.). None of these relationships were

significant.

I ran chi squares to see the relationship between whether the fourth comments were

positive or negative and feminine touch and gender poster (X2 (6, N = 160) = 2.652, p = .851.),

ritualization of subordination and gender poster (X2 (6, N = 160) = 4.664, p = .588.), licensed

withdrawal and gender poster (X2 (6, N = 160) = 4.455, p = .615.), kissing pout and gender

poster (X2 (8, N = 160) = 1.244, p = .996.), body display and gender poster (X2 (10, N = 160) =

15.720, p = .108.), makeup and gender poster (X2 (6, N = 160) = 3.182, p = .786.), muscle

presentation and gender poster (X2 (10, N = 160) = 10.989, p = .358.), and setting and gender

poster (X2 (6, N = 160) = 2.652, p = .851.). None of these relationships were significant.

Then, I ran chi squares to see the relationship between whether the fifth comments were

positive or negative and feminine touch and gender poster (X2 (6, N = 153) = 6.364, p = .384.),

ritualization of subordination and gender poster (X2 (6, N = 153) = 4.646, p = .590.), licensed

withdrawal and gender poster (X2 (6, N = 153) = 5.064, p = .536.), faceless portrayal and gender

poster (X2 (6, N = 153) = 13.635, p = .034.), kissing pout and gender poster (X2 (8, N = 153) =

8.971, p = .345.), body display and gender poster (X2 (10, N = 153) = 3.233, p = .975.), makeup

and gender poster (X2 (6, N = 153) = 1.398, p = .966.), muscle presentation and gender poster

(X2 (10, N = 153) = 3.820, p = .955.), and setting and gender poster (X2 (10, N = 153) = 3.091,

p = .979.). None of these relationships were significant.

I ran chi squares to see the relationship between gender of the poster and whether

comments were positive or negative. These relationships weren’t significant for the first (X2 (2,

N = 184) = 3.699, p = .157.), third (X2 (2, N = 165) = 3.008, p = .222.), fourth (X2 (2, N = 160)

= .266, p = .875.), and fifth comments (X2 (2, N = 153) = .094, p = .954.). I also ran chi squares
Gender and Instagram 75

to see the relationship between race or ethnicity of the poster and whether comments were

positive or negative. These relationships weren’t significant for the first (X2 (8, N = 191) =

9.558, p = .297.), second (X2 (8, N = 179) = 2.195, p = .974.), third (X2 (8, N = 172) = 8.024, p

= .431.), fourth (X2 (8, N = 166) = 10.647, p = .223.), and fifth comments (X2 (8, N = 159) =

11.725, p = .164.).

I ran chi squares to see the relationship between race or ethnicity of the poster and

whether or not comments were about gender. These relationships weren’t significant for the first

(X2 (8, N = 208) = 8.845, p = .356.), second (X2 (8, N = 205) = 12.301, p = .138.), third (X2 (8,

N = 203) = 7.766, p = .457.), fourth (X2 (8, N = 200) = 4.991, p = .759.), or fifth comments (X2

(8, N = 198) = 4.046, p = .853.). I also ran chi squares to see the relationships between the

gender of the commenter and whether or not comments were about gender. These relationships

weren’t significant for the first (X2 (2, N = 191) = 4.989, p = .288.), second (X2 (2, N = 179) =

1.104, p = .576.), third (X2 (2, N = 172) = 1.878, p = .391.), fourth (X2 (2, N = 166) = 0.49, p =

.976.), and fifth comments (X2 (2, N = 159) = .671, p = .715.).

I ran chi squares to see the relationship between whether or not comments were about

gender, the gender or the poster, and gender posing. These relationships with the first comments

and feminine touch (X2 (6, N = 200) = 9.834, p = .132.), ritualization of subordination (X2 (6, N

= 200) = 11.118, p = .084.), faceless portrayal (X2 (6, N = 200) = 7.818, p = .252.), kissing pout

(X2 (8, N = 200) = 10.701, p = .219.), body display (X2 (10, N = 200) = 9.435, p = .491.),

makeup (X2 (6, N = 200) = 9.062, p = .170.), and muscle presentation (X2 (10, N = 200) =

17.345, p = .067.) weren’t significant. The relationships with second comments and feminine

touch (X2 (6, N = 197) = 8.031, p = .236.), licensed withdrawal (X2 (6, N = 197) = 7.716, p =

.260.), faceless portrayal (X2 (6, N = 197) = 6.621, p = .357.), kissing pout (X2 (8, N = 197) =
Gender and Instagram 76

7.373, p = .497.), body display (X2 (10, N = 197) = 11.508, p = .319.), makeup (X2 (6, N = 197)

= 5.925, p = .432.), muscle presentation (X2 (10, N = 197) = 8.137, p = .615.), and setting (X2

(10, N = 197) = 16.023, p = .099.) weren’t significant. The relationships with third comments

and feminine touch (X2 (6, N = 195) = 4.554, p = .602.), ritualization of subordination (X2 (6, N

= 195) = 6.047, p = .418.), licensed withdrawal (X2 (6, N = 195) = .629, p = .996.), faceless

portrayal (X2 (6, N = 195) = 4.928, p = .553.), kissing pout (X2 (8, N = 195) = 8.499, p = .386.),

body display (X2 (10, N = 195) = 6.098, p = .807.), makeup (X2 (6, N = 195) = 4.993, p = .545.),

muscle presentation (X2 (10, N = 195) = 8.392, p = .591.), and setting (X2 (10, N = 195) =

7.791, p = .649.) weren’t significant. The relationships with fourth comments and feminine touch

(X2 (6, N = 192) = 3.151, p = .790.), ritualization of subordination (X2 (6, N = 192) = 5.193, p =

.519.), licensed withdrawal (X2 (6, N = 192) = 5.076, p = .534.), faceless portrayal (X2 (6, N =

192) = 5.195, p = .519.), kissing pout (X2 (8, N = 192) = 7.258, p = .509.), body display (X2 (10,

N = 192) = 8.241, p = .605.), makeup (X2 (6, N = 192) = 4.083, p = .665.), muscle presentation

(X2 (10, N = 192) = 6.019, p = .814.), and setting (X2 (10, N = 192) = 9.410, p = .494.) weren’t

significant. The relationships with the fifth comments and feminine touch (X2 (6, N = 190) =

6.155, p = .406.), ritualization of subordination (X2 (6, N = 190) = 5.431, p = .490.), licensed

withdrawal (X2 (6, N = 190) = 1.770, p = .940.), faceless portrayal (X2 (6, N = 190) = 5.525, p =

.478.), kissing pout (X2 (8, N = 190) = 4.234, p = .835.), body display (X2 (10, N = 190) =

5.682, p = .841.), makeup (X2 (6, N = 190) = 3.324, p = .767.), muscle presentation (X2 (10, N =

190) = 8.188, p = .610.), and setting (X2 (10, N = 190) = 9.269, p = .507.) weren’t significant.

I ran chi squares to see the relationship between the gender of the commenter, the gender

of the poster, and whether comments were about gender. This wasn’t significant for the third (X2

(4, N = 140) = 4.384, p = .357.), fourth (X2 (3, N = 133) = .627, p = .890.), or fifth comments
Gender and Instagram 77

(X2 (3, N = 125) =.882, p = .830.). I also ran chi squares to see the relationship between the

gender of the commenter, the gender of the poster, and whether comments were negative or

positive. This wasn’t significant for the first (X2 (3, N = 146) = 2.617, p = .455.), second (X2 (8,

N = 140) = 14.270, p = .075.), fourth (X2 (6, N = 133) =4.246, p = .643.), or fifth comments (X2

(6, N = 125) = 4.038, p = .672.).

You might also like