16. Data Science & Presentation Skills - A Talk With Margot Gerritsen (Women in Data Science Worldwide & Stanford University)

Listen and get insights into Data Science & Presentation Skills in this talk with Margot Gerritsen, (Founder & Former Executive Director Women in Data Science (WiDS) Worldwide, and Professor [Emerita] Stanford University).

We talk about: presentation framing and motivation;  understanding your audience and the 2 - 3 messages you want them to remember; similarities between your presentation and writing a newspaper article; ending your presentation with a memorable takeaway; explaining where your data is coming from and how you look at the data; overcoming your fear of public speaking - you have nothing to lose, but you have something to gain; being honest about your data and the outcomes; practicing your presentation - the value of multiple dry runs; managing stage fright and anxiety; reducing presentation fears by inviting criticism early on - it will make you a better presenter. 

Guest Info

• LinkedIn: Margot Gerritsen
• LinkedIn: Women in Data Science

Resources

• Book: Lies, Damn Lies, and Statistics (Michael Wheeler)
• Website: Women in Data Science
• YouTube: Women in Data Science

Support the Show: Hey There! Become a supporter and help us create great Women in Big Data content for listeners everywhere. It starts at $3 per month, and you can cancel anytime

Mentoring Program - Women in Big Data
Mentoring is essential to success at every stage of a women’s career, both as a mentee and mentor. The many WiBD mentoring programs are open to WiBD members and cover opportunities for junior, mid-career, and senior women in technology. Not yet a member? No worries. By joining a mentoring program, you automatically become a WiBD member. Both membership and mentoring are free of charge.

Website: Women in Big Data Podcast
LinkedIn:  Follow - Women in Big Data
Contact us: datawomen@protonmail.com

Transcript

[00:00:00] Intro: Hey, Hello! Welcome to the Women in Big Data podcast, where we talk about big data analytics and career topics. We do this to connect, engage, grow, and champion the success of women in big data. 

[00:00:18] Margot: "When I started in data science and I started presenting data, I tried to be really aware of how I was presenting it and explain why I was presenting data a certain way. And it's been part of my life ever since. Every time I talk about data, I try to create for the audience a real good sense of the framing that I'm using and motivation for that framing." 

[00:00:44] Desiree: In this episode, we talk with Margot Gerritsen about data science and presentation skills. Margot is the Founder and Former Executive Director of Women in Data Science Worldwide and Professor [Emerita] at Stanford University. She's passionate about supporting women and other underrepresented groups in science, technology, engineering, and mathematics. 

Let's start!

[00:01:12] Desiree: Margot, welcome to the Women in Big Data Podcast. In this episode, we're going to talk about data science and presentation skills and why it is important to have these skills. So, Margot, you presented a lot of complex data science concepts. How has this influenced your public speaking and communication approach?

[00:01:33] Margot: I cannot even begin to describe how important it's been. Let me first say thanks for having me on the podcast. Going back to presenting, one of the interesting things I've learned over time is that data for those who are in it - and working with data - is relatively easy to grasp. And you know, after many years in the field, you play with data, it becomes so second nature to you. That's what I mean to say. 

So, understanding what kind of lens you put on when you're presenting data, whether to look at the data in absolute values or to look at the data in relative senses, all that just comes naturally to you after a while. But for most people out there, who are not working with data every day, data actually often present a lot of challenges. They find it hard to understand statistics. They find it hard to interpret data. They get bombarded with data every day of their lives through the news, social media. And very often, data is presented in ways that are not conducive to really understanding: it's confusing. 

And at the same time, when people hear data, they hear a number. They feel that it must be associated with fact, right? So when we talk, we say: well, we are going to be objective; we're going to present you with facts; and here are the facts, are the data. And so, it's really strange for a lot of people. I think they hear data, they think, 'Oh, these data, they must be true.' And yet, at the same time, they're often talked about in such obscure, difficult, and contradictory ways that it makes it very hard for people to understand them. 

When I started in data science and I started presenting data, I tried to be really aware of how I was presenting it and explain why I was presenting data a certain way. And it's been part of my life ever since. Every time I talk about data, I try to create for the audience a real good sense of the framing that I'm using and motivation for that framing.

[00:03:55] Desiree: Okay. And by framing, do you mean you use a kind of framework that came naturally to you? Because I can understand when you're analyzing a lot of data that you already get in your mind a kind of story.

[00:04:10] Margot: Yeah. So that's the problem, right? When we, as data scientists, look at data, we always approach the data from our own framing. So, we look at the data with a particular purpose. We may look at some data and not at other data. We may look at data in a particular way. We may look at the simplest way to explain it. Sometimes, you look at absolute data; sometimes, you look at data relative to something else. So, you may look at absolute numbers or numbers per capita, for example. And for us, it's very clear. And we look at the data with our own lens on it because we're trying to understand a certain situation. We're trying to solve a certain problem.

And what we often don't do as presenters is explain that frame to others. Say: hey, I'm looking at this data from this or that perspective; this is what I'm trying to achieve; and that is why I presented data in that way. Really important to explain: this is where I'm coming from; this is how the data relates to the problem I'm trying to solve; this is the lens that I'm using to look at those data; and this is why that lens is important to me. 

And it's also important for us to understand that people may come into this discussion with a different lens. This is where often the misunderstandings happen. People come in with a different lens, and if you don't explain that right from the get-go, you may get misunderstandings. You will get arguments. You realize, 'Oh my goodness, we're just looking at this from two totally different sides, both legitimate.' But we should understand this and explain it. 

[00:05:50] Desiree: So, that also means when you are doing the analysis, you are really into it, but then, actually, you have to step back and see: how can I present this to an audience so that they understand what I'm telling and where I would like to go.

[00:06:05] Margot: And for that, you also need to understand your audience. So, every time I give a presentation, think about, you know, what will the audience know? What would the audience probably like to get out of my talk? What sort of data would intrigue them? How should I present these data to appeal to the audience? To appeal like that is of interest to them and that they understand, and we can communicate about them because, of course, the end goal of any presentation is not just to show what you've done but to get people interested in using what you may have found out and then take it further.

We're always about advancing. This is the start of a dialogue and the start of further investigation. And so for that reason, it's so important to frame it well right from the get-go. And I've certainly given talks and then the end of the talk saying: oh no, this was a completely different audience, or they misinterpreted what I said.

And when people write about data - be that reporters in newspapers, or people writing journal publications, people writing books - of course, a lot of mistakes are made. And the first time that really came to my attention was by reading some books written around this concept of Lies, Damn Lies, and Statistics, where brilliant authors were looking at the ways that statistics are misrepresented, miscalculated, and misinterpreted.

I talked earlier about how it's so hard for people often when they're not in data to really understand it. And statistics is one of the most misunderstood. And you see this in daily life also all the time. You see it in the newspaper, but you see it also in people's behavior. I have these discussions all the time about statistics in daily life. For example, risk estimation is one of the hardest things for people for that same reason. 

[00:07:57] Desiree: If I listen well to you, then you do use a kind of framework, although it comes more natural to you. 

What I always do when I have a presentation, I visualize three pillars. It's the beginning, the middle, the end. And in the beginning, you tell: what am I going to tell you, and why is it important? And then, in the middle, I say: how can we reach that? We have a kind of strategy or some examples. And then, in the end, is a kind of call to action. So, and there also comes in, of course, who is the audience? What do they know? What is of interest to them? How can you create some hooks? So, that's how I prepare for a presentation. And I always say: keep it super simple. Don't make it complicated because then you often get lost, and also the people to whom you are presenting get lost. 

[00:08:48] Margot: You're so right. I always think about beforehand what are the two or three messages that I would like the audience to remember?

And when I help others refine their presentation, I always ask that before they start to present: what are the three main messages you'd like to convey? And then later, we discuss: what did you actually convey to them; did they come out okay; how did you weave your story around them? And that's true for every single part of a talk as well, right?

So, if you just think about it, in our typical talks where we use presentation slides, then for every slide it's important to ask yourself: is this slide necessary; is this slide conveying what I really wanted to convey; is it helping me or not; why do I have that slide; what is the main message of that slide; does the slide convey that main message really well? 

And I often liken a presentation to writing a newspaper article. If you look at a newspaper article and the way that good articles are set up, then first of all, you know, the title is really intriguing. And that's like the start of a presentation. The first thing you read must intrigue you. The first thing you say must make the audience want to stay and listen to you. Just like the first thing you read in a newspaper article must convince me as reader: hey, I want to read this article. 

Then the first paragraph of a newspaper article really gives you sort of the general gist of this article. And it conveys the motivation often why this article is written. And it lifts the veil just a little bit. You want to know more about this. So that's that first paragraph. And then every paragraph after that, it's like a slide in your presentation. And when you look at a good newspaper article paragraph, the first sentence of that paragraph has the essence of it. That's the main message. Because if you speed read a well written newspaper article, often you only need to read the first paragraph. And then the first sentence of every following paragraph, and then you read the end paragraph again, which has this sort of takeaway or amazing conclusion. And so you can do the same with slides.

The title of every slide is like the first sentence of a newspaper article paragraph. That is your biggest real estate, right, on the slides. That gets the most attention. And it's often by people used in a way that's not very interesting. So, how many times do you not go to a presentation where the slide titles say things like introduction, or step one, or conclusions? There's nothing interesting in that slide title. So that's where your main message goes, just like in the first sentence of this newspaper article. And then you fill in the rest through your presentation. But it forces you to think about, okay, what's the main message here? What is the flow of my talk?

And then, at the very start, do I lift enough of the veil to interest people, and do I frame this talk correctly? So, I really like how you've done it, because you always want to end your presentation with something that people can take away. And very often when people present, what do they use the last minute for? Thank yous. So, they even have a slide that says: thank you. It takes all the energy out of the room when you say that, you know, short of closing off the discussion. And then, there are all these thank yous to people that the audience probably don't even know.

So, I always say to people and help them present also for our WIDS conferences: end with something that the audience can mull over. Something provocative maybe, or a call to action could be something that loops back to the beginning. It could be a call to action, it could be what can you do with this stuff now, or what you should be focusing on, or it could end with a little bit of excitement about what still needs to be researched also, right? So, it could say: well, we've done this, we made a little contribution here, but this amazing problem is still outstanding. I always talk about outstanding questions. Wouldn't it be great to work on that together? That could be a provocative ending. And then I think you have a really good talk, but it takes some time.

And then going back to data - because we're talking about the data - if the data are a main part of your talk, if your talk has been inspired by data, if you're conveying knowledge about data, the data are sort of the main character in your story. Then, it's so important to introduce them early, to explain where they come from, to frame how you look at the data.

And the other thing that's always really, really, important to tell the audience too, is not just how you look at the data, but where this data comes from. And very often that's neglected. So, data is presented as facts and say: here's the data, it's wonderful; it helps me complete this picture. But is the data really complete? Was the data collected by you? In what way? Did you have a certain bias already when you collected data? If it was collected by other people, why did they collect the data? What sort of bias may have come in there? Is the data actually complete? Well, never really is. But how incomplete is it? Can the data be erroneous? All of that is really important, and very few people do this. They just say: here is a description of the data that we looked at, and here are our results; and here's the outcome; and that it may be a data mining problem, maybe data that was used to train a model - and then they're talking about the results of the training.

But very seldom do you hear people talk about the data first. What is this? How should we interpret this? How good is it? Who collected it and why? And then how did I work with the data? Did potential buyers come in, and so on? So, there are so many interesting aspects to talking about data and these are all really good things to think about.

[00:14:36] Desiree: Well, the way you talk about it sounds very interesting. easy, but I know when you get all these questions, you can be overwhelmed and even sometimes have fear to present, especially in the beginning of your career, you can be overwhelmed. And what do I have to say? And how do I look? Et cetera. What do you recommend to people who are having fear for public speaking?

[00:14:57] Margot: It is so common. And for many years on Campus also, I taught presentation skills to students in my institute, and the first thing we did is just present often, because that fear goes away by doing it more often. The more you do it, the better. So we would have meetings where everyone presented every time we met.

Sometimes we just present the same thing over and over because it wasn't what you were presenting, it was the fact that you were in front of an audience that freaked most people out. Also, teaching really helps, and I've always liked to teach, and when you're teaching there's some of that same trepidation.

You're in front of an audience, they're paying attention to you. Very often, actually, when people first start out, and it was the same for me, you're so self aware when you start teaching or you start presenting, you believe the audience watches every part of you. They're scrutinizing you. They reflect in their body language how they feel about you.

So it's all about you. You're too self aware. So when I started teaching, when I started presenting, if someone in the audience yawned or looked bored or turned away from the talk, I thought it was about me. And this is the fear that a lot of people have. Now, of course, after a while you realize, no, it's not about you, particularly not when you're teaching.

I mean, it may be in the morning and they may be angry because they didn't sleep very well that night, or. They may have just received an email that upset them, they may be really stressed because they have to give a presentation later on. People live in their own world and you're just such a small part of it.

So, in fact, in a presentation and in a lecture, you often have to fight to They keep people's attention. It's the opposite of what you're fearful of. What you're fearful of is that they're constantly scrutinizing you, that they pay you too much attention. What you really should be fearful of is that they're not giving you any attention whatsoever and they're not listening to you.

So that's the first thing that you have to realize. What I always tell the students is you've nothing to lose, but you've got something to gain. The other thing is when I ask my students, what are you most worried about apart from the judgment that you think is coming your way? What's really difficult is that they believe that even though they're often the experts, that there will be people in the audience who will see right through them and will find all the mistakes and will ask them about them.

The first times I presented at conferences as a student, I was so scared that someone would realize halfway through my talk that what I was doing was just highly flawed. They would not just realize this, but also tell the audience about it. And they would ask me questions at the end that I just could not answer.

And the truth of the matter is that if you give a talk about something, you're the master of that talk. You are probably the most knowledgeable person in this room, is your stuff that you're presenting. It's a different thing when you have to present for a team and you're not really on top of everything.

But most of our presentations, we're pretty much on top. Do we realize that we don't know everything about this topic? Of course. Of course. The stuff you know you don't know is always larger than the stuff you know you know. So we always come in with a healthy dose of self criticism. But the audience knows a lot less than you.

You're going to give them something, right? They're going to come out of this talk richer. And questions that are asked are often relatively benign, or they may come across as critical, but it's often because either you didn't frame it well enough so people are confused, or the people listening don't quite get it.

Grasp what you're saying the first time around and need to think about it. And these questions can be interpreted as clarifying questions. So questions are never that scary. Now, it is true that there's a lot of competition, and I've certainly been at conferences where the questions asked, were not asked in a very friendly way.

in a very accepting way because where people may be feeling competitive with me. And then academic culture can be brutal and the culture in the industry can be brutal. So there may be people in the audience who believe that they can raise their own standing by putting you down. And that I think has for me always been the hardest to help others with because there is that risk.

And particularly when you're different, you are scrutinized more because you stand out. You are different. So people are going to pay attention and people in the audience may not be comfortable with you in that role. So they may look for ways to put you down and that is a completely different ballgame.

I've certainly had this happen to me in conferences, and I've never accepted that behavior from others. I would try to stay calm and collected, and then go and talk to a person afterwards, and try to sort this out. But it can be difficult. 

[00:20:18] Desiree: Yeah, I absolutely agree. The way I handle that, and I also had to learn that, is I prepare very well.

And then it doesn't matter what other people think. I know what I want to say. I know where I want to go. I know there will be people in the audience who will understand it and the people who are critical. They are welcome. Maybe I can learn from it. And otherwise I have to go on because it's not important.

[00:20:44] Margot: Yeah. And that's very good. And you learn it by having. experiences like this. The first time it happens, it has an incredibly big impact on you. The first time it happened, I feared the consequences. I thought, Oh dear, now this person thinks so negatively of me. That person had. some FaceTime in front of this audience, they listen to him, they're now all biased against me.

But what you quickly find out is that the audience is not. They often see through this also. They understand the behavior and the people even talk about it sometimes say, Oh, yeah, no, that's that behavior again. And they don't actually like the attacker for it either. But it's tough to stand your ground and it's tough to stay calm and collected.

Later in my career, after 20 years of giving talks, when you're in a conference and you see people line up behind a microphone to ask questions, you can already tell from their body language what they may say. And the other thing that I notice is that I try when I give a presentation not to boast about what I'm doing.

I'm not making myself bigger than I am. I'm honest about what I know and what I think could be interesting to the audience and also about what I don't know. And I see this with students also, that when they're attacked, often the response is, well, next time I'm just only going to talk about the good things I've done.

I want to make myself bigger because I'm being attacked. I'm going to inflate myself a little bit. Whereas the best action is just to be quiet. As you said, completely honest, said this is what I know, this is what I don't know. No, this thing is not the best thing since sliced bread, as so many people say.

This is not the algorithm to end all algorithms. This is not the approach in machine learning or computational mathematics that makes all of you redundant. This is not better than what you're doing. This is not applicable to everything. This is something that works here, that seems to hold promise. And I know, and I understand the boundaries, it doesn't work there.

And when you present in that way, then people also don't feel threatened by you, right? And then maybe you inspire someone else to also not overinflate their findings, because there is a lot of hype, and that makes it harder often for an audience to To really understand, you know, I believe people more, and I tend to listen better to people who are very open about shortcomings of algorithms, potential problems these algorithms may cause, messiness of the underlining data about uncertainty that they've introduced.

And if people say, hey, we tried to break this, and here is what we did to try to break this. Here is where we've seen our own boundaries and where we should still be better. But very often when you read in social media or you listen to talks, people don't try to break their own algorithms. They are only showing you cases where the algorithm works miraculously well, and they're not open about it.

All the times where it didn't actually work all that well and particularly in data science is so important to be honest about this. If you have all this hype about research and the data that you've been using and the outcomes of your data algorithms and you make them all look so much better and you filter everything out that's imperfect, then when people actually start working with it or use it later, then um, the disappointment comes.

And I think then you're at a real risk of losing some of your reputation. So it's just better to be honest. So like with everything, there's so many parts of life where we completely accept and we know that without practice, you never get there, but somehow In many areas of academia and also industry, people are supposed to do this really well, very fast, and then they're surprised when it doesn't work.And it's the same thing with presenting, with playing with data. You don't really get those insights instantaneously. I have students in classes who come to me after taking just a few lectures and say, I can't do this. I don't have what it takes to do this class. I said, well, you've only participated in three lectures.

You've only done maybe one assignment. This is way too early. It takes a lot of practice and you have to do it again and again. Just like with running or playing tennis for me as presenter and also as mentor. One of the most positive experiences of my life is this women in data science conference that we have every year at Stanford.

I work with the speakers. And so we say: hey, you have 10 minutes and it would be great if you'd get it done in 7, because then we have a gap. The reason for it is we're on the time schedule with live streaming. It's like a television show. You have to run on time. And the first time we practice, they never manage.

And then we practice again. And we talk about it and we go through slides and we practice again and then they practice themselves. They do dry runs themselves. Then we do another dry run and then they come on stage and every single one of them nails it. And that's just because they've practiced and they often say, wow, I've never practiced as much and now I see practice makes perfect.

And people always ask me, you know, after giving so many presentations and being so in front of an audience so often, do I get stage fright? Am I nervous? And the answer is absolutely every single time. And every time I start teaching a new course and have a new audience that I need to get to know, that needs to get to know me, I get stage fright and I have anxiety.

I still have them 40 years later, right? And so that's completely normal. And I also think a little anxiety can help because a little bit of adrenaline going through my body actually helps sharpen my mind. So I think it's the same for athletes. When they do a race, they have raced hundreds of times in their life, but they're still nervous.

And I think it helps them sharpen. So if you feel nervous, often when people present and they feel nervous, they think the nervousness is a reflection of their inability. And I tell you, it's not the case. It's just natural. If you're in a research group, if you're in a team and you feel that presenting makes you nervous and you try to avoid it, don't avoid it.

Seek it out. The more you do it, the more comfortable you be with it. Discomfort is not something you can outgrow by avoiding. It's not that 10 years from now, all of a sudden you're not nervous anymore. You have to outgrow this by doing and the earlier you do it, the better. So start presenting in relatively safe environments, volunteer to present in your team, in your study group, in your classroom with a teacher that you trust.

Volunteer to present at conferences, then go into mini symposia at conferences that are often smaller audiences with people that working in your area, maybe understand you a bit faster, a little bit better, and go from there. Take every opportunity. And then for every time you give a presentation, do dry runs.

I do a lot of walking up and down in my room and pretending I'm on stage and just doing real dry runs in my office out loud. 

[00:28:23] Desiree: Well, I think you gave a lot of insights and a lot of good advice, so we are coming already to the end of the podcast, but is there anything else that you would like to share? 

[00:28:33] Margot: If you're really, really worried about a presentation because you're worried about your work being exposed, then think about how can you reduce those fears by inviting criticism, commentary early on.

[00:28:50] Desiree: That's a nice way of saying, inviting it. Then you don't have to fear it.

Margot:
Yeah, because if you've been avoiding that in work, then of course giving a public talk is going to be a nightmare. But if you, in your work, invite people to look at what you're doing and you take their criticism, their constructive criticism, hopefully, to heart, then The confidence grows and the fear diminishes and that makes presenting easier because you've already tested it on people.

It will make you more comfortable in what you're doing. It will make you not shy away from presentations. It will make you give more presentations. It will make you a better presenter. 

[00:29:37] Desiree: Wonderful. Thank you very much for this, Margot. It was really a pleasure to talk with you. You're welcome.

[00:29:46] Outro: Thanks for listening to the Women in Big Data podcast. For more information and episodes, subscribe to the show or contact us via datawomen@protonmail. com.

Tune in next time!