Talk data to me: How to present youR data to any audience (Freda Warner, CIHR) | posit::conf(2025)
Talk data to me: How to present youR data to any audience Speaker(s): Freda Warner Abstract: Translating data and analytics to diverse audiences is a vital part of any data scientists’ job, no matter what level. Strong data communication skills are valuable for every shiny app or quarto-generated presentation you give, and involve more than just avoiding pie charts! This presentation will cover creating a narrative with your data, identifying key messages, creating effective slides and visuals, and delivering an oral presentation that is not only articulate, but will help your audience understand and remember your message. posit::conf(2025) Subscribe to posit::conf updates: https://posit.co/about/subscription-management/
image: thumbnail.jpg
Transcript#
This transcript was generated automatically and may contain errors.
Hi everyone, I'm Freda, and I'm an epidemiologist turned data scientist. And this means that I've kind of pivoted from conducting research with neurological data to analyzing health research funding data at the Canadian Institutes of Health Research.
This also means that throughout my career, I've kind of consistently been presenting various types of data findings to various types of non-data audiences. And these audiences have ranged from basic scientists to clinicians or students or patients, and now government executives and policy makers. And I've learned a lot, and I don't think I need to convince anyone here of how important it is for us to be able to effectively present about our data.
Without this skill, our data and our findings stop with us and can't move forward to have any further impact. So today I want to talk a little bit about how we can improve our data presentations beyond just avoiding pie charts.
And how many of you have ever gone to a presentation, and you walked away and you were like, wow, that was such a great topic, but I had such a hard time listening and focusing, and I don't really remember anything. Hopefully that hasn't happened this week. But no matter what, I might argue that you're not the problem, because I'm constantly re-inspired by this topic.
Just earlier this year at work, I was assigned to help a team with some of their analytics, which is pretty common. So we worked together, we did a lot of analyses, and when it came time to present the results to their stakeholders, they asked me if I wanted to join the meeting, just to see and also maybe be there in case there are any data-specific questions. And in their meeting, they talked about aims and goals and questions they wanted to answer. They presented a bunch of results. And the speakers were good, and the slides were very nice. But even though I was the one who had actually conducted all of the analyses, I still managed to walk away from that meeting and not really remember anything and not know what their main message was.
They had all the pieces that were there, but they hadn't been connected or presented in a way that made them stick with me. It was kind of like we were walking through this forest of random facts and findings, but there was no signpost or leader to get us to this destination, or in this case, our main message.
And just like you would never start leading a group on a hike without first knowing where you were going, your destination, you should never start working on your data presentations without first identifying what your main message is. Because once you have this message, you can identify what you need in order to lead your audience towards it in a way that gets them to listen and understand and remember what you're saying.
And just like you would never start leading a group on a hike without first knowing where you were going, your destination, you should never start working on your data presentations without first identifying what your main message is.
And this is not always easy when you're presenting a lot of numbers to non-data audiences. Your presentation probably has a bunch of different types of data and analyses, maybe some Shiny apps. But if you're not able to identify what your main message is, your audience definitely won't be able to either. And we're often so excited to show a lot of stuff. But we know that our audiences are never going to remember everything we show them. And so kind of prioritizing this one thing we can get to stick, that's our goal here.
Identifying your main message
And a lot of speakers this week have done something similar. They'll be talking and they'll say, if you're only going to take away one thing from this presentation, and they'll just go out and say it. And that's really the idea here.
And unlike this other story where it was about another group, I've also, of course, been guilty of kind of poor communication in my presentations. One particularly vivid memory for me was very early in my career it was my thesis proposal. So it was a very important day. And I was presenting to a very smart and important group. And I just got mixed up as to what the goal of my presentation was. Instead of identifying a message and kind of telling a story, I just wanted to show them everything I'd done. I wanted to show them all my data and all my analyses and that I'd left no stone unturned.
And ultimately I wanted to impress them. And I wanted them to think I was smart. And I thought maybe if I provided so much information, they'd be less likely to ask a lot of questions at the end because questions are scary. And as you can imagine, my presentation was just like this busy, incoherent mess. And I was trying to include so much information that I talked way too fast. And this all came together and resulted in a ton of questions because I couldn't identify my message. They couldn't identify my message. They didn't know what my proposal was about. I wasn't leading anyone anywhere. I was just kind of presenting defensively instead.
Because once we identify what our main message is, we can then identify what information we need to lead our audiences towards it and avoid getting lost kind of in this forest. And unlike that thesis example that I was just saying, you don't want to try and cram in all your data and all your analyses into your presentation just in case. Because this usually results in our message getting lost. And it's not that you're hiding anything from your audience, or I hope you're not, but in focusing on what's relevant, you're getting your message across, and they're always welcome to ask about that extra stuff later.
Just like your presentation is kind of one big message, you should be focusing on kind of one piece of evidence and one smaller message at a time to keep your presentation clear and concise. Because sometimes less is more.
One of the ways I was able to kind of wean myself off this need to overfill and overcomplicate my presentation was to have supplementary slides, sometimes a lot of supplementary slides. Because often my need to overfill came from either being so excited to show everything, you know, what if they were interested in it, or God forbid, what if it came up in the questions and I wasn't prepared. So having all the supplementary slides really helped me to focus on my story, knowing that they were there in case I needed them. Kind of like, you know, your first aid kit in your backpack when you're hiking. Good to know it's there, hopefully you won't need it. Usually you don't.
Using visuals to support your message
And once you have your main message, and you know the evidence you need to support it, you can start incorporating visuals into your presentation. And visuals are our tools. Sometimes they're figures, sometimes they're images, sometimes they're just text. And I'm not going to sit here and be like, you know, this data should be with this kind of figure, you should use this plot for this kind of data, because our work and our preferences can dictate that. I mean, I was requested to make a pie chart this year. It's fine.
What I more want to focus on is that our visuals are there to support us. And one of the ways our visuals can help support us is by making sure kind of each of these smaller pieces of evidence and these smaller messages that we're using, we're focusing and emphasizing on one at a time.
And one recommendation I had, again, early on in my career was this five by five rule. Some of you might have heard of it, some of you might have not. But basically it just means no more than five bullet points per slide, no more than five words per bullet point. And I don't like hard rules. There are exceptions to everything, including this.
But I do think that this kind of rule helps to remind us about two important things with regards to our visuals. And that is, like I mentioned, our visuals are there to support us. Our visuals are not there to do the heavy lifting. And you should never be able to send your slide deck to someone and have them understand what it's about without your accompanying presentation.
The second thing is that our visuals should be kept as simple as possible. And I don't mean the figures themselves, because sometimes you're not going to be able to have a simple figure. You need a complicated one. But if you find yourself trying to include a lot of figures on one slide or trying to cram in more and more text, you should probably be breaking this slide up into multiple slides. One, because you're probably not focusing and emphasizing on one message at a time. And two, because this just helps our audiences with short attention spans pay attention. Just like our presentation is kind of one overarching message, each slide should have one concluding thought.
I remember a visiting scholar one time coming to do a presentation at my old work. And he was supposed to give a 20-minute presentation. And he went up to the front, and he plugged in his laptop. And some people noticed that he had something like 180 slides, and they just panicked. But his presentation, his visuals were supportive, and they were simple. It's just he had a lot of kind of repetitions where he emphasized different areas or highlighted results, and he had a ton of supplementary slides. And his presentation was really good. Because we shouldn't be afraid of too many slides. We should be afraid of really busy slides.
And I think we can all know when we've, our visuals have become kind of a hindrance instead of helpful. Because you pull up your slide, and you start talking, and at some point you look at your audience, and they're peering at your slide, and they're kind of reading it, and they're trying to understand it, and they've completely stopped listening to anything you're saying, which might be happening right now. Because you've become secondary to your visual.
And even though figures are often preferred over kind of big tables or big blocks of text to keep your audience's attention, what's more important than the type of visual you use is that you're walking your audience through what it is and what it means. This visual, for example, is fine. This is on our website. You can find it, and it also has the accompanying data that was used to make it.
And the data table has funding type rows and fiscal year columns, and it's just a bunch of numbers. And the figure is a much clearer way to show these patterns over time. But the second I bring it up, I should still be kind of leading audiences through my presentation. I say, you know, here we have figure three, CIHR. Again, that's my work, Canadian Institutes of Health Research. CIHR fiscal year investments by funding type. And then I kind of dive right in. I say, you know, on the y-axis here we have investments in millions of dollars, and on the x-axis we have fiscal years. And then you go into the data, and you tell them what the message you want them to take away is.
It can be as simple as, as you can see here, investigator-initiated research and research in priority areas have been consistently increasing since 2000. This helps the audience to stay on track with your narrative instead of just kind of throwing them a compass and trusting that they can find their way. It's even better if you can really emphasize the area that you want them to pay attention to.
And lots of times I've been presenting to audiences that I thought were very smart, I assumed much smarter than I was, and it made me really shrink away from leading. I thought I shouldn't over-explain things that I might look kind of silly if I was talking down to the audience. But you know your data, you've done the analyses, and this is your presentation that you're supposed to be leading. Even though it can be really tempting to assume that any figure you present is so obvious, that the message is so obvious, they haven't been looking at it for as long as you have. And no matter what, you should be walking them through it the way I did, kind of axes to data to message.
Similarly, if there's a figure you're not gonna walk through, don't just throw it on your slide just in case. That's when our audiences start doing the peering and wandering away from us, and that's when our message gets lost. You are there to lead, and your visuals are there to support you.
Delivering your oral presentation
And finally, you've done your visuals, you can start working on your oral presentation. And your oral presentation job is that final step to really get your audience to listen to and understand your message, and hopefully to get that message remembered. And if there's only one thing you can recommend, or I can recommend, about the oral aspect of your presentation, it's to say it out loud at least once, ideally many, many times.
We already spoke about articulation. They really force you to say it out loud before you come here. And as you're making your slides, you're probably already kind of starting crafting what you're gonna say in your head. Maybe you're making some slide notes. And I know that when I start making slide notes, I just tend to write in this weirdly formal way. It's how I write, it's how I email. I don't know if it's me or if it was academia. But the second I start saying it out loud, I'm able to identify what doesn't actually sound natural or feel natural for me, and the areas that I trip over. And also if I have an area without kind of a main message, I kind of trail off without a concluding thought and it helps me to highlight those.
They say you shouldn't write a script for your presentation, and I agree, but the more times you're able to practice saying it out loud, the more natural it'll become, both in the way you present it, but also if there's little hiccups like technical difficulties or a phone call, or a random question that comes up. Practicing it out loud is so important.
Another important aspect of your oral presentation, in addition to kind of getting your message across, is to really create an environment that encourages audience to interact with what you're saying. And again, I touched on how questions are terrifying, and I'm often really nervous about them in my data presentations. But an old supervisor of mine used to say that questions are a compliment, and this is because you want your audience to both have understood, but also care enough about what you're saying to bother asking a question about it. And even if they ask those really hard questions, that can help maybe identify areas that we've missed before it's too late, or even kind of indicate where we should be heading to next. Questions are a good thing.
But an old supervisor of mine used to say that questions are a compliment, and this is because you want your audience to both have understood, but also care enough about what you're saying to bother asking a question about it.
And in order for an audience to ask questions, you of course, the most important thing is that they understood what you're saying. But equally, providing this kind of encouraging environment where they're comfortable interacting with you is just as important. And that's kind of done in ways that we've already touched on, walking through your findings, language like, see here, what this is showing, how to get your message across, versus, you know, obviously this, assuming they can interpret everything, flipping through slides, that can make your audience a little too lost, or even just too intimidated to ask any questions at the end.
And this is related to the age-old recommendation of knowing your audience. While you might be comfortable just providing a bunch of statistical outputs to a group of data analysts, or scientists, or whatnot, you probably want to provide more context and more explanation to a more diverse group of students or stakeholders. Even a posit, this is my first time here, I assumed they'd be, you know, pretty, we'd all be on the same page, but once you start talking to people, you realize how much diversity there is, even within an audience like this.
So although you should be able to tailor your presentation to the audience at hand, generally keeping things clearer, a little bit simpler, just expanding more in your explanations, pausing for your messages, that'll really help get your message across, and encourage your audience to interact with your findings, which is what we want to invite.
Another tiny note about inviting audience interaction, I'm sure everyone's familiar with this, say you went to a presentation and it's got a lot of results, a lot of different findings, and someone asks a question, they're like, oh, can you go back to when you were talking about this, and they're like, no, no, no, like, further, further, further, no, it was before, you know, the slide, the blue dots and the red lines, it can be any presentation, but especially when you have really content-heavy presentations, just number your slides. It can help save time and just, you know, one little extra step in inviting that audience interaction we're looking for.
Bringing it all together
And now you've done it. You've translated your findings. You identified your main message, you used data to create a clear path towards that message, you incorporated your supporting visuals, and you delivered an oral presentation that led your audience step by step through the journey that you wanted them to take. In keeping your message clear and understandable, you've allowed your findings to move beyond you and have impacts like, you know, inspiring further research or informing policy or improving business operations.
We need to be able to translate our findings to a variety of audiences because in doing this, our findings and our data doesn't just stop with us. It can have further impact. And in order to do this, we need to learn to lead our audience towards the message and not leave them lost in the data.
And as a reminder, you can submit questions via Slido. We have a few here. Do you suggest different slide decks for presenting versus sharing out? So if we're presenting more slides that tell a story but condensing into fewer slides for sharing out?
No, I still don't think fewer slides is the answer. I still think when you're sharing something externally, you will need to write more. You're not going to do the five-by-five rule, but I think the main goal is to still focus on that one message at a time, and that doesn't mean you need to condense things. That's the best way to get your story across, one small message at a time in order to tell that story. I don't think condensing really serves us even when we're sharing out, except for people being a little intimidated, but that's okay.
How do you suggest organizing the supplementary slides to provide all the extras without the live presentation? That's a great question, and I think this is really a preference thing. What I would recommend that I do is when I have my supplementary slides, I have these big title blocks separating them to sections, so when someone asks about fiscal outcomes, say, okay, and I can find the title slide and then look within that section. I tend to organize them in what I think is most likely to come up, but I think the way you organize them beyond that is up to you.
Someone asked, I love this concept. What's your advice for organizations where the data's secondary message and supports an idea to the decision pipeline, where they tend to have more words on slides rather than figures and tables?
If the data's secondary to the message, then it's just one of your chapters, right? It's just your whole presentation is not about the data, but rather in each of your main messages, your data's only going to be kind of one of the little slides or one of the little chapters in that story, and that's okay, too.
How can we utilize LLMs to improve our own conversation tactics tied to specific audiences? You know, this is probably not a great question for me at the moment. One thing that often makes, if we're trying to explain something complicated, and you've probably noticed this in a lot of the talks that have been going on today, if you can kind of incorporate a metaphor into your story, that really helps people to listen and kind of connect with it and remember it, so LLMs can be great maybe for asking or getting a story across, but again, not really my strength, I'm afraid.
What do you think about slides as documents, slide decks that contain details that could be consumed later? Oh, right, like again, sending, yeah. Instead of a lot of presentations. I mean, again, you are going to need to include more information. I really want to emphasize, I don't think you should be condensing into fewer slides. I think there's going to be a lot of slides, and there's going to be more text, but still doable, which is not my favorite.
You have beautiful visuals that really added to your presentation. What practical tips or tools do you recommend? Thank you. I was really nervous about the visuals because with our articulation coach, I didn't have them ready when I was supposed to. I mean, I don't want to sound like a broken record, but these, I really worry that these were too simple, that they were too kind of bare bones, and it did make my presentation harder to remember, but I've received a lot of great feedback about them, and I'm just really hammering home that the simpler you can make them, the more the audience is going to remember them. I do, you know, little things that I didn't bring up, these particular slides take into account color blindness. That's a nice thing to do as well, but supportive and simple is really the message there.
Great, and last question. Do you make any supplementary docs for your presentations since you're not designing the slides to be self-explanatory? How do you handle requests for slides? Absolutely. If I'm sending my slides to someone, I would definitely send supplementary information if it's something really content-heavy. You know, things that you refer to in your presentation, you can then make notes, see supplementary table, that's a really good tactic for doing this sort of thing. I agree with that, yes.
