Sports bras vs. lifting weights: The stereotyping of gender in sports recommendations

 

An exciting and important new research is on the way! Gender equality has been a much-debated topic for a long time. With the rise of social media, this debate has been taken online and has become even larger in the new, ubiquitous world of the digital age. Social media have made a great difference in the positions of those affected by gender stereotyping, but it might also work the other way around. By studying people’s online behavior, companies can develop strategies to reach different target groups. The online behaviour of males and females seems to be very different and it is not unthinkable that companies actually want to reinforce the stereotyping of male and female, in order to increase their revenue. For this project, we would like to use Facebook and investigate to what degree this platform allows companies to influence males and females. If we take sports as our area of focus, how different will the content be that men and women see? Do men really see more “manly” content on their timeline, such as articles with tips about weightlifting? Is the sports content shown to women more concerned with something regarded as mainly female, such as fashion or recipes? And how about news outlets, how differently are they portraying themselves to men and women?

This issue is worth writing about because it is dealing with certain types of stereotypes, at least that is what we are expecting. We assume that people of a different gender, in our case male and female, are being treated differently on Facebook (and on every social media platform for that matter) in terms of what is being recommended to them. We expect that Facebook’s recommendations are based on gender stereotypes to give people content that they want to see. But who says, for example, that a man and a woman should get different content recommended to them when they are both interested in sports? When both a man and a female like ‘super bowl’ related content, should a guy get articles about the football game and a girl blog posts about what designer J-Lo was wearing during the halftime show, for example?  We live in a time where topics such as feminism, equality in terms of gender, LGBTQ, etc. are very upcoming and important, and are being adhered to by a large part of the world population. Because nowadays it is impossible to imagine a world without social media platforms (such as Facebook), it is important for these platforms to adjust their behavior regarding these crucial topics. However, first, it is up to us to investigate the extent to which Facebook is adapting their recommendations to the type of gender behind the profile.

We decided to use Facebook Tracking Exposed to help us gather the data that we need for our project. As one can see in their privacy statement, Facebook Tracking Exposed is a browser extension that gathers data only from public posts to help you understand Facebook’s algorithms. On the website, it is also mentioned that in order to make sure that your privacy is being protected, Facebook Tracking Exposed makes sure that you are aware of what you are doing when you install the extension to your web browser. And, the privacy statement of Facebook Tracking Exposed also states that the browser extension only assembles public posts and if you do not want the browser extension to collect your posts you just have to change the settings on your posts from public to private. Facebook Tracking Exposed is also very transparent in their privacy statement about what information they collect, so people using the browser extension know exactly what they can expect. With the help of Facebook Tracking Exposed, we want to collect data from two personal Facebook accounts and from two ‘bot’ accounts. We want to compare female users to male users. All accounts will be following the same sports pages and we would like to see if the accounts get the same posts on their newsfeed or if there might be some differences between the content that the accounts can see on their newsfeed. Comparing two ‘bot’ accounts with two personal accounts might lead to interesting results since the personal accounts have a much longer history than the ‘bot’ accounts.

We looked at some existing statistics and numbers when it comes to gender stereotyping in recommendations and in sports. The first interesting article we found was from the Soccer Betting 236 website. This article shows the difference between women and men when it comes to sports in many different fields. A good example is that with boxing there are 17.981 professional male boxers but only 1.228 female boxers. The difference with tennis players is much less severe: 1.966 male tennis players versus 1.235 female tennis players. When you look at the number of female amateur football players in the Netherlands you can, however, see that it’s never above 25,5% of the total football players. All these big but also smaller differences in professional and amateur athletes could stereotype sports with genders. It doesn’t mean a sport is male or female but algorithms could be influenced by this. It’s not easy to gain good insight in how the platform’s targeting algorithm works, however, since Facebook is very cautious of what they share about the working of their algorithms; very little information on their preferences logic has been made available to the general public. But what do we know? Facebook says they will show their users what’s most relevant to them, but what does their algorithm black box consider relevant? How sensitive are their algorithms to stereotyping? Well, very, according to research from experts on human-computer interaction. Their experiments show that, for example, posts that would stereotypically be more interesting to men (e.g. bodybuilding), can deliver to over 80% men, and the posts that would stereotypically be of the most interest to women (e.g. ballet dancing), can reach to over 90% women. Biased outcomes? Sounds like it.

The interview

For our interview, we will interview two active Facebook users that are interested in sports and at least like some sport-related pages. One male user and one female user. We will interview one of them in person and with the other over the phone/Skype. We’ll focus the two interviews on their experience with the Facebook algorithm and the recommendations and posts they get on their feed. Next to their own experience, we will find out what their opinion of this whole phenomenon is. It is important to know what the users find of this recommendation system based on multiple factors and especially based on gender. What is their view on this? Because maybe they don’t mind their personal algorithm or maybe they do, like with the increasing attention to gender neutrality. We’ll introduce the two Facebook users for you below.

The female user that we will interview is Dionne van Lint. She is a 19-year-old UvA student, she used to play hockey and now likes to do some fitness (sometimes alone, sometimes with a friend). We are going to interview because she likes to do sports but is not necessarily interested in following them (online). We want to know if despite this she has noticed some things around gender on Facebook, around sports or maybe something completely different. We think this might be interesting because she is not particularly following many sports teams, but just because of this she might even be better to tell if she has (unconsciously maybe) noticed stereotyped recommended content on Facebook.

The male user that we will interview is Dylan Wassink. He is 25 years old and a singer-songwriter who likes to do fitness. Just like Dionne, he is not following many sports teams, so this might be good for comparing the two. Also because he is a singer-songwriter his feed may be less biased since his type of music (Ed Sheeran etc) is liked by both genders, as opposed to if he would be interested in rock music for example. Due to logistical reasons the interview will take place on the phone and will be recorded.

The questions we will ask are all important, but they won’t all be asked if the person that we interview answers it in another/different question. We’ll focus on their experiences and opinions.

  1. Gender is a big thing nowadays, like gender neutrality or gender-focused marketing. And this also occurs on social media. Do you notice this yourself?
  2. And what do you think of this whole phenomenon?
  3. Then, how much do you use Facebook daily yourself?
  4. Any other social media you use?
  5. Do you notice the algorithm on FB?
  6. Do you think the algorithm and recommendations get influenced by gender?
  7. Do you see this yourself?
  8. Do you have examples of things you noticed that was recommended to you in relation to your gender?
  9. Do you follow any sports pages?
  10. If so, do you see these sport recommendations sometimes targeted to male/female users? Like a fitness post with tips in how to get a very masculine/feminine body.
  11. As a male/female do you sometimes see sports posts with a team/person that is your own gender?
  12. Do you see your own gender more or not? And could you elaborate?
  13. What do you think of the whole idea that Facebook uses gender in recommendations?
  14. Some people see it as a good thing because you get recommendations that are probably more in line with the individual users, do you agree?
  15. Does it scare you that Facebook uses gender as a factor, or is that not a problem at all?
  16. Generally, what is your whole opinion about the subject we talked about? So gender in sports recommendations?
  17. Finally, does this idea that gender influences recommendations change your mind of Facebook as a social medium or not?
About

We are Joris Binsbergen, Fenna van Dijk, Lynn Hoekstra, Karina Strauch and Anna-Lisa Vuijk.

Tagged with: , , , , ,
Posted in Data Journalism 2020