Login

Data is Female, with Sian Davies | The Ironhack Podcast

On the Ironhack Podcast, Sian Davies, Lead Data Teacher @ Ironhack Barcelona, tells us about gender bias from a data perspective.

You’re reading an interview with Sian Davies, straight from The Ironhack Podcast. Every week our hosts, Tim and Dan, catch up with Ironhack alumni, teachers, and other tech professionals for industry deep dives and personal success stories.

To listen to this episode (or to binge-listen to every single one, we wouldn’t judge you!) check us out on Spotify, or wherever you find your favorite podcasts.

Sian Davies is a Data Analytics Lead Teacher in Ironhack Barcelona, previously in Berlin. She started Data Is Female, a quarterly meetup in Barcelona, designed to give a platform to women in the world of Data to talk about their experiences and share their knowledge. Sian shares her insights with us, from a Data Scientist perspective, on gender biases in the Tech and Data industry, and how we can overcome them, within ourselves and in society. 

Teaching the Ironhack Data Analytics Bootcamp

A: My name is Sian Davies. I am British, as you can tell from the accent, I'm 38 years old, and my pronouns are she/her, And I am the Lead Data Teacher for the Ironhack Data Analysis Bootcamp in Barcelona.

Q: We first met when you were a Data Analytics Teacher in Berlin.

A: There was an opportunity to come over here and teach, and the timing was better, and the weather was better, so I decided to give it a chance. I'm teaching in person. That's the main important difference. Back in Berlin, we only managed to teach in-person for a couple of weeks before the pandemic kind of ruined that party. I was teaching online for the best part of 8-9 months. Now I get to teach in my own classroom in person in Barcelona, which is day and night compared to that. It's just a different experience, so much more human, and you can really see if the students are struggling, and feel that energy in the room and react and respond, and we're having a much better time. I'm really enjoying it.

Not to say that remote doesn’t have a bunch of positives. I think it can work really well; we adapt very well to the remote format and there's great advantages for the students. They can watch the recordings back, they never get caught behind. We have this breakout room concept, so they get a bit of privacy to discuss away from the teacher. There's wonderful things about it! And they are learning to work remotely, which is probably going to be their future job environment as well. But from the teaching staff perspective, I think we've been forced to see the best of it and coming back to the classroom reminded me why I love teaching.

Q: Well, it does feel like lockdown is looming over us again, so let's not speak too soon.

A: No, absolutely. And at the beginning of this move to Barcelona, I said I’d be very surprised if we get to graduation without going back on into remote format. So I'm enjoying it. I'm enjoying every week.

Q: In the long run as well though, it's probably a good thing because the students get to experience both modes in the hybrid format.

A: Absolutely. And the ideal teaching situation is that the students get to meet each other in person, and also bond with you and the TA. And then if you do have to go remote, it happens later in the course. That works much better that way than trying to do the whole course remotely, or starting remotely and then trying to retroactively bond. It’s like dating online for six months and then seeing if you actually get along in person.

Q: The last time we spoke, about a year ago you were teaching in Berlin. Since then, has the curriculum changed?

A: Yeah, absolutely. The curriculum at Ironhack is always evolving. Each teacher, me included, adds to the materials and improves them. We actually introduced some new stuff today that we've never done before, and I wrote the materials only last week and asked my students for feedback on it. I think it has to change with the times, the needs of the class, and the motivation of the teacher.

Q: I know our Web Dev bootcamp changes all the time, so I was wondering how fast it moves in Data Analytics.

A: I would say at least the languages are not changing as fast. You add details and maybe the libraries change, but we're still using the same core technologies as we were back in December. But we are trying to adapt to changing requirements in the job market and new technologies, and I also adapt to what I know about my students. I get opportunities as well, for example, I was recently given access to a bunch of training materials for Databricks, as I'm collaborating with them, and I use it in the classroom.

Q: What's Databricks?

A: Databricks is a cloud platform for cloud warehousing but also joining the blocks together. Hence, bricks. So you're joining together the various parts of your remote workflow in order to build a machine learning project or do some data analysis and reporting, and dashboarding whichever tools you need to use. It's an enhanced way of joining all that together, but in the cloud; and we have nice prebuilt pipelines between the tools, which saves time since you don’t have to sit and code to connect those pipelines.

Q: Cool. That's one of the lessons I've had about Data Analytics: a big part of it is actually presenting the data. Gathering it and wrangling it is one thing, but then presenting it and getting your insights across…

A: Yes, visualizing it and constructing a narrative doesn’t matter if you can't get someone to listen to you, believe you, and take that insight on board. The presentation is two things: the narrative, and also the visuals. Unfortunately, it's not where we spend the vast majority of our time. We spend the most time getting the data, cleaning it, preparing it, and shaping it.

Q: Like when in Dev we spend most of our time finding the single piece of syntax we missed or spelling mistake we made, rather than actually building the cool project.

Data Is Female: Why We Need a Platform for Women in Data

Q: The main topic we wanted to discuss in the show is this network or community that you are building in Barcelona, called Data Is Female.

A: We had our first event last Friday, and I want to do it every three months if I can. The idea is to, just as you said, start building a network here in Barcelona. It's something I've always wanted to do. I've pushed myself a little bit into doing it because, looking around my classroom, we've got three amazing female students in the class. And then the rest of my students, eleven of them, are all men.

Q: I think that Data is predominantly male.

A: The whole industry is male dominated, especially Data Science, and there’s a number of structural reasons for that, but we can do something to address that balance and give women a platform. Because that’s another thing I find frustrating. As a woman, I certainly feel shy when I go out and I meet people that I don't know, but if somebody asks me to do a speech, show up at an event, or do any kind of representation, like a podcast or an article, I'm like, yeah, absolutely, let's do it! Even if I’m a bit unsure of how it will come out. But I find that’s uncommon among women. You ask any guy to appear on a podcast or talk to your students and they’re up for it even if they're not that highly qualified. But ask much more qualified women, and she’ll hesitate and question whether she has anything to say or anyone will want to listen to her. We need to fight that in ourselves, and I figure I can help a little bit with that.

Q: The idea of being intimidated if you’re not naturally extroverted keeps coming up in the podcast episodes; if you graduate from a bootcamp along with hundreds of other students, how do you actually rise above the noise?

A: Having the confidence to get out there and say something that may not be groundbreaking and may not be the top of the top is a gift, and I think that the more platforms we can create for women where they feel like anyone can go out there and have something to say, the more they can talk about their experience. And they can talk about anything, but they definitely have stuff to say: they are working in Data! They are working in tech, but they're, for some reason, a bit reticent about it.

Q: Congratulations on, on starting the group. So Data Is Female aims to happen every three months in Barcelona, on campus: is it streamed online?

A: We didn't stream the last one online because it was the first one, but I think it's something that we could do in the future. I really wanted to see if we could try to build an in-person community, but, as you said before, we don't know if there's more restrictions coming. So it makes sense to do it in a hybrid format as much as possible. We’ll probably get more guests as well if we can do it online. Let's see what happens. Let's try and build that network.

Q: Why do you think that it is intimidating for women to speak up in the world of data? Is there an equal amount of interest in the field from women and men?

A:Well, if you look at the statistics about schools, it’s tied up with the STEM problem; I think there's a lot of overlap. STEM is an acronym for Science, Technology, Engineering and Mathematics. It seems that these are the technical subjects that girls are very keen at and do really well at until they turn 16. And then there's a sudden drop. They just drop out of it, and for a very long time we didn’t know why. Lots of organizations are now doing stuff about this, like Women Who Code, and I have a friend, Debbie Forster, who is involved at the UK Tech Talent Charter, which is fantastic. Many organizations are working on diversity and inclusion and making sure that girls have access to STEM and that they're excited about it..

However, there's still a massive imbalance to address there. One thing is the dropout problem, since studying science and mathematics obviously takes people down the road of data. But then there's also this concept of who is a gamer and spends lots of time on video games, which is another avenue that could take women towards tech and data, yet is still very much seen amongst young people as a male pursuit and very unusual for women.

Q: Actually, I read about the history of this male/female divide within tech, which was very different in the earlier stages, right?

A: Yeah. María shared this great slide about how many women were involved in technology and computing in Europe and the United States until the eighties or so, when computers started being seen as crucial to our future development and economic generation. And then suddenly women were sidelined, and very deliberately so, by organizations, by governments. They were told: this isn’t a place for you anymore, and we were quite happy to have you work in computing during the war when we needed you, but now it's actually become quite important, so go off and do something else while the men take over.

I was also reading about child psychology, and maybe you guys can align with me on this. As a kid, I was told that I had to be perfect and get things right, and that I had to look good, be tidy and tidy up after myself, never get messy or mess up. Whereas boys are encouraged to break things, and try, and fail, and get dirty. Because Data Science and Machine Learning (and coding in general!) are all about getting things wrong to then learn how to do things better, I think this during very early socialization we're giving girls the message that they can't do trial and error.

Q: The generation that we all grew up in was very differently raised. We’ve seen this recently emerging trend of raising children in a gender-neutral fashion or allowing them to break gender stereotypes, though.

A: That will change things, yeah. Now kids are being encouraged to code regardless of gender and I think there is a generational change coming, but at the moment we still have a real lack of women in the tech industry, especially at the top. We don't have great role models to follow. So yes, there are a lot of systemic problems, but there's a lot of hope as well.

Q: I think this conversation about early conditioning is really interesting and complicates the conversation around diversity, because it’s elementary that there should be equality, but the inequality is much more deeply-entrenched and much more complex than that.

I used to have the approach of simply treating people equally and thought that would be enough, but the longer I spend in tech, the more I realize that's not true at all. The disparity is still huge, and the general attitude towards women in tech seems almost… it's very strange until you see it.

A: I felt similarly, that as long as I was doing my job and I was there and visible, I’d be a leader, a role model, and I could go on pretending that the world was gender blind to me just because nobody has explicitly stood in my way. And when they have been critical, sexist to me, spoken over me or when I see anything directly, I call it out and I always have done, and I've always got away with that. And I'm very lucky because I'm white and I'm privileged and I can leave a job and find another. But you're not really doing much to help anyone else. It’s a bit like recycling: you convince yourself that that's all you need to do and the planet will be saved.

Q: But, and I wonder if it's the same with the diversity issue, it's very difficult to see the benefit that you alone are making. Like 71% of all pollution is generated by a hundred companies in the world. It’s way bigger than that. However, it is gaining more momentum.

A: I think you have to see our actions around both of these issues as ripples in the pond. You don’t necessarily see them, but there are ripples nonetheless.

Gender Biases From a Data Analytics Perspective

Q: At the event you mentioned the margins of difference being small yet significant, whereas in other contexts huge margins can be considered very close calls. So I wondered, in the world of data, how do you call biases and see these trends?

A: It's difficult to see the trends when they're marginal. This goes back to some of the core concepts; we teach about confidence intervals and margins of error. And you have to assume that data is noisy, it's one of the laws of big numbers. The more data you collect, the closer the totals are gonna be split down the middle; elections and electoral polls are a classic example of that.

Even if the margins are tiny, they can be statistically significant, and there are methods to find out if they are. But the biggest problem with when you've got quite small differences is your own bias when you encounter patterns and gaps; confirmation bias will make you interpret the tiniest gaps as proof if you don’t do statistically rigorous testing to make sure.

Q: Is it a useful exercise then to give yourself deliberate, false bias on things?

A: I think it's a very useful exercise to define your hypothesis before you start. What we tend to do is define a null hypothesis, and you start from the basis that there is nothing there, rather than starting thinking that you have something. Writing your alternative hypothesis also helps you to see your own bias; you refer back to that question and check yourself.

Q: To bring it back to the women in data topic and unconscious biases: before we started, you mentioned this fascinating article, a data analysis of how gender bias impacts GitHub contributions.

A: I was looking for interesting examples of men versus women in different scenarios, and I found this article where they found a remarkable difference in the amount of successful pull requests between male and female GitHub members. And for example, 25% of women had nearly 100% of their pull requests accepted, whereas for men it was closer to 13%. And that is a substantial difference. That's not marginal. I discussed it with a couple of people after the event. However, the article still shows how women’s contributions are less accepted and there is a bias against women there, still.

Q: It might go back to what you said about men being more confident. 

A: Yes and, and this is a known problem with women. They’re generally not as confident putting themselves forward unless they know for sure that they have every qualification. And I’m doing exactly what I said, I’m looking for evidence to back up my supposition that women are scared to be anything less than perfect, but I do think there's enough repetitive or repeated evidence there to suggest that there is something going on. It feels wrong, and we can certainly do something about it.

Q: Another one of your guests actually mentioned patience. It can take a long time to get where you want to be, and patience is a great tool, but if you are aiming for perfection, that's an unachievable goal.

A: It doesn't matter how patient you are. You're never gonna get there if you try to catch up with the technology, because there are new models and languages coming out all the time. You don’t have to be 100% qualified for the job. That is actually how my Ironhack interview went. Men apply for jobs when they know they’re not quite there, and women should as well.

This is not a women-only event, everyone is welcome, men and non-binary people too. But I want to give women a stage and a chance to talk about their experience and to share their ideas. We do need to take more opportunities. 

It might take a while for you to be ready in this industry. You might have the job success really quickly, but maybe you are not ready and that's okay. Take that job if it serves you as a stepping stone, and if you change your mind later, after you’ve learned more and figured things out, you can switch. Sometimes you just need to have that keep your eyes open and look for those opportunities, and be willing to learn.

Q: Well, congratulations for finding a career that you are happy in. Sometimes you have to take a look around and acknowledge that you’re doing great.

A: Yes, I think you have to every now and again. I try to get my teaching assistant to do this sometimes. We have hard weeks, we have great weeks, we have great days in hard weeks… and sometimes you just have to give yourself a pat on the back. Because we have to celebrate those things.

If you want to know more about Sian’s career in Data and her Data Is Female initiative, you can listen to the full Ironhack Podcast episode here.

Sian told us about how Data Analytics can reveal statistically vital information and teach us more about our unconscious biases. If this piqued your interest, take a look at our Data Analytics bootcamp, and start your career in Tech!

Are you enjoying this article? Keep learning about Data Analytics!

Take the first step into tech and find out more about our Data Analytics bootcamp

Related blog posts about Outcomes

What Is Ethical Hacking?

Read more...

How To Pivot To A New Career (And Get Into Tech!)

Read more...

What Makes an Ironhacker

Read more...

Life After Ironhack - 18 Months On, With Kara Payne | The Ironhack Podcast

Read more...

Building a UX Portfolio to Get Hired, With Sarah Doody | The Ironhack Podcast

Read more...

Becoming a Portfolio Professional, with Ben Legg | The Ironhack Podcast

Read more...
Stay up to date on our latest news and events. Sign up now!
Please type your name
Type your last name
The email is not valid. Please try again

By applying I accept the Privacy Policy and the Terms of Use