Are TED Talks contributing to reinforce gender biases?

As a woman studying data analytics, I was pretty surprised and upset when I noticed something the past days: when I think about certain occupations/professions, I immediately think of men.

I went to google and typed some of these professions on “images search”, and the first pictures that appeared to me were also of men. Take a look:

So I started researching unconscious gender biases, which is defined as unintentional and automatic mental associations, stemming from traditions, norms, values, culture and/or experience. It basically made me realize: “well, of course I pictured men in those occupations, this is what we usually see on the TV, google, podcasts, movies…”

The thing is, this kind of bias is not harmless. Research shows that this is one of the reasons women have a harder time getting leadership positions, and it is also a reason why “masculine fields” stay that way. To change that, we need to give voice to women in these professions: we need to make female scientists our role models, we need to give voice to women in tech, we need to see female bankers thriving.

So I decided to analyze what I think can be an important tool for that change: TED Talks. TED Talks are a space where people from various fields of expertise communicate to the general public, with a great reach potential. TED was conceived 37 years ago, and because of its huge success, it is now in several countries, and has subtitles in more than 100 languages. The number of times TED Talks have been watched is in the billions, and they have more than 19 million subscribers in YouTube.

Taking that into account, I analyzed the representation of women across different topics, to see if TED was helping to enforce, or to break our gender biases. It is important to say that the analysis comprehends the entire database from TED Talks given in english, since 1987 until 2020, accounting for more than 4000 Talks.

For the purpose of the analysis, I had to create 2 new variables in my dataset: gender, and general topic.

To be able to determine the gender of the speaker, I considered only the name of the main speaker (some Talks have more than one), and used a Python package called “gender-guesser”. This package uses a very big list of names across different countries, and by the person’s first name, categorizes it into “male”, “female”, or “neutral”. Since “neutral” comprehended less than 10% of my database, I worked only with names that could be identified as “male” or “female”.

I also created groups of subjects, like technology, science, business, culture, and communication. In total, there are 14 groups. Each Talk is classified and labeled according to the topics it addresses, and it usually has more than one label. It means one Talk can address “business and culture”, and the other can address “business and technology”, for example.

For the analysis I used Python, mainly Numpy and Pandas. I also used Tableau to make the data visualization easier and faster. Now I’ll get into some findings:

I measured overall representation of male and female speakers in the database, and we have 37% of women, and 67% of men.

Below you can see the topics each gender addresses the most:

Men talk more about technology, science, and culture, while women’s top 3 are culture, science, and psychology. It is important to take into consideration the fact that there are 67% of men in the sample, and 37% of women. So the graphs do not have the same scale, and for any topic that you look for, there is going to be a bigger male representation.

But there is another perspective for us to see, and that is the evolution of representation of men and women as TED speakers over time:

As we notice, the huge difference we have on the representation of genders in the database is mainly because of the years before 2017, which means TED Talks improved a lot when it comes to giving voice to women.

But this is not enough for working against gender biases: these women must be talking about subjects and topics that we usually attribute to men. In the graph below you can have a better idea on how this goes:

Fortunately, we can see that the representation in fields like technology, economics and business is getting more equal in the last years.

By the data analyzed, we can conclude that in the current period, TED Talks do help us to work towards the goal of breaking gender biases. By giving women space to talk about their expertise in economics, science, and technology, we begin to shift the “male image” that these words and professional fields tend to have.

I know it is only the world of TED Talks, but it is a movement of change. And it is only starting. All of us have a lot to gain by promoting diversity into companies and industries, so let’s keep this in mind while fighting for a more productive and equal society.

Background in business and marketing, also telling stories with data