Diedre is a Lead Data Storytelling Trainer at StoryIQ, where she helps organizations improve their communication with and about data. She helps professionals embrace their inner data translator to engage more stakeholders with data-based decision making. An accidental math teacher, Diedre learned the power of demystifying numbers in New York City classrooms and the power of influencing decision-makers with data during her time running WeTeachNYC.org for the NYC Department of Education. Diedre is an Adjunct Lecturer at Hunter College in New York and has spoken at NCTM, iNACOL, and Learning Forward about adult learning methodology and best practices in professional learning.View the profile
About the talk
Data is everywhere – which makes fluency in understanding, using, and communicating with, and about, data an essential skill. Often, we focus on building impressive visualizations and models only to find our audience confused about what we need them actually to learn or act on, from our data. The most impressive data analysis is useless without the ability to communicate essential takeaways and offer up persuasive recommendations clearly. Learn how taking a storytelling approach to sharing your data can help get your work noticed, and recommendations heard.
Thank you so much for joining me today at the end of a long day filled with some very heavy material machine learning. I'm going to give you a little bit of a break from that and talk about the essential role that ba analytics translator plays in Modern Workforce and how adopting that role can help you really get your work noticed and help you become more of an influencer through your work. In our brief time together today. I wanted to talk about why data can't always be our first priority and understanding why
we need to reconsider our audience and design with our audience in mind. Then we'll take a look at how we live action through translating analysis and making sure that we're really in mind and then I was just making small changes to invite your audience in to your expertise to not leave them out there. Wondering what in the world you're talking about. A few changes can really help you build a much larger conversation on people in your conversation and really
help connect with staplers for Simpson Q&A at the end. If our goal is to have a dialogue with stakeholders and decision-makers, we can't start with the data and part of that is because not everyone speaks data. So, you know, just think about your experience the past couple of days last week about dinner with someone who doesn't speak that language fluently that went dead making those sounds from your desk. Housing data is a challenge and no one who is in a leadership position in an organization
likes to feel what they don't get something but it often happens with me. I'm just showing you what it's like for someone who doesn't speak data to wander into a conversation with you and your teeth. . This is super cute. If he's never watched the entire like 2 minute clip babies have a conversation in their own language that no one else understands. They're having a great time. They're laughing their telling jokes right for the adults who are filming
this have no idea what they're saying. This is the position we unfortunately and unintentionally what are stakeholders in when we just start spewing data language and using our jargon, we have a conversation or not, even a conversation piece would have watch what's going on without being able to interact and our goal is to change that By really taking on the role of being a data translator. And that's really what we wanted him to be. And before I go any deeper into this talk. I
didn't feel like I should come clean to you. I would not consider myself a date assigned. I am feeling a fast in R and python. I'm pretty sure even if you gave me $1,000 I could not accurately tell you what that is, but I am a numbers person and that means but I craved understand data and it's patterns, but I'm not an expert in creating visualization for complicated programs that we need to keep analysis. I am a DJ Tiesto and dedicated need a translator and that makes me a valuable part of the organization because that means I learn to
speak both languages the language which I respect and I respect everyone who's doing very technical work with it, and it doesn't understand that. I'm going to stop talking to one another with me or you prescribed a real importance to sharing data. Why don't you to preach to you, but that's not our bosses and our team needs to sort of cringe when we start talking about data your partner or spouse for your roommate really doesn't care about what you did all day because it bores them. We need to think about ourselves as David translators
if we want to be understood and celebrated by those people who are not in the room with us because those people are often our audience and the decision-makers who can help push our work forward, but we need to do without simplify the complex in ways that would work with data or dumbing down what we learn from data. We need to translate our work into something that is Meaningful and understandable to are less data from and colleagues. I've been 10 years Jane high school math in New York City. And in this role. I learned very very quickly
that is fluency and math and numbers. Did not equip me to be a math teacher. In fact, my numeracy actually need it made me very difficult teacher to learn from because I didn't realize that I needed to interpret language. That was so understandable to me into something. My students can understand and that's a position for many of us find ourselves. When we shared a strip of metal translate. Do you need for a person to understand? Our work is the first step in being able to move your work from informing to influencing?
This guy here for many many years and finally realize is that decision-makers didn't want me to just go run off as interesting as that was for me school relationships that I could find. That's not what they were looking for. They wanted me to really sit down with them understand their business challenges and then go back to the data to find a way to inform possible solutions to the challenges. Why so many gated projects fail to live up to expectations? Because we don't do this
piece of stopping and thinking about our stakeholders and decision-makers is that we dive head-first into the data looking for something interesting to us. Unfortunately, that's not the piece that always carries value. Understanding what your decision makers Comfort level is with data with tools with jargon is the most importance and knowing what level of Translation your work needs. And what we don't want to do is make any assumptions underlying knowledge of a process or program that our
audience may have and we want to be crystal clear by the impact and results of our analysis. Not sure why our analysis important and what we can do with it and by making the decisions about how to have this conversation to stakeholders, you're going to eat your audience's understanding and by engaging them in a conversation that makes sense and that they can be an equal partner in the first A heart that might be your approach to solving problems the way you approach a problem when you approach going into a big data at
and make all of the difference in translating your results. And I'm the first to admit that I love trying a small or try to find a small obscure interesting elevation in a big date of that takes time and it focuses on again what we're interested in rather than trying to consider the problem, but this data can help us all for this challenge that I can help us address to take a look at how focusing on the analysis and focusing on the results and not the Sterly our audience
can sometimes getting our rent. By looking at what is an almost daily challenge in my house at what to have for dinner. So I have been right and I I wish we were better meal planners. We're not it often comes the end of the evening of each other and what we going to eat tonight. And so I am in this case is my stapler my decision maker life under he's the one who really needs to make this decision about dinner year than I am. Just giving you these parameters hear that whatever we be delicious and reasonably
priced. Okay. Well, there's a lot that I could do with that information. Thanks pretty broad problem. So in order to help me figure out how to share the possibilities make a decision and a week into a final decision so we could cook something or someone else could cook and I'm going to make the executive decision that we go with someone else Cooks not feeling it right now. So mad at someone else's cooking that leaves me to options we order in or we die now.
I live in New York City. We just open restaurants up inside and it's getting a little cold outside went to the idea of dining out is not that appealing to me. Okay. Well this problem space reduce the number of options I have I think you had how long it takes for the food to be delivered. Is it 30 minutes or less than 30 minutes and knowing my audience is Ryan is already a little bit in doubt by this whole process and putting together to help us make a database decision about dinner. I know that he starts to get cranky
the laundry goes in that dinner. So I'm going to pick the quickest delivery time possible and would inevitably comes down to a New York is doing a pizza or pizza for dinner. And this case I'm going to go with pizza. All of that work for the narrow down where I need to focus my search for data on a possible solutions going to bring value to my stakeholder food quickly that we don't have to cook. But still allowing me to jump into the data to help make a decision. At this point Ryan is rolling his eyes at me. Right?
He just wants to order pizza and say no we have to make it the best decision possible about the piece of meat and so I go to Uber Eats and I just grab some dinner real quick for the local pizza places that will deliver in under 3 minutes staff and you're working. Then again to a little cleaning convert that and so now I've got my data and I need to go into analysis form. So thinking about what I need here going in at a total cost for how much does it to Pieces is going to cost us in the delivery fee
may be getting an average of numbers that compares choice to and I like to feel pretty certain that these the ratings were looking at her after it. So I just decided to have dinner. And now I've done my analysis and I bring this to Ryan who is my stakeholder and say here you decide what do we have for dinner? Here's all the information at your fingertips. What do you think this reaction is? We are in his eyes nowhere closer to ordering dinner than when we started. And in fact, you just waited for me to validate this database
approach a problem. I haven't translated my analysis into something that's easy for him to use to make a decision. I haven't embraced. What being did a translator means in the last step of trying to share my insights. No one needs to or wants to really figure this out basement table. So I'm going to give him a little bit of residual and then make a scatter plot here comparing the average star rating from ubereats to how much it cost and Brian said it had to be delicious and reasonably priced. So I'm definitely going to
get rid of my upper left quadrant four things are more expensive than average is less delicious beverage that helps need a decision. Tell Nick and Ryan work just enough time for this is getting angry right this whole analysis is pretty weird to think the best choices and give him a couple options of Y and it might just be adding a little narrative to help him understand why I've made these choices. So we have our table maybe you thought that as you're looking at the table as well or tacos for dinner and for some reason but that does
lactase key Norseman. I'm here. Ryan can visualize all the possibilities and rationalize why these are the options. I'm getting to him. That's what we do is translating. We have a central role of making sure our audience can use our analysis to make a decision. And so we want to move people to action and we're concerned about bigger problems than pizza. I'm sitting here working bigger data sets. That approach of sharing our results in a way that's easy to understand is the
essential component of being an analytics translator and that's how you can help make sure that you and your stakeholders are speaking the same language. Tell me the share just a couple of other ways for you to consider your analysis and visualization to help you move towards being a more common language for your eyes. The first is considered the fact that your analysis is interesting. Not necessarily to everyone. So you might do an analysis and you come up after doing a random Forest dalasis.
There's Your Meme decrease after seeing you here. Now, you know what we need to do if your audience is perhaps the manager of a bank right Bank probably doesn't know you supposed to have no idea how to read this increase accuracy or what the impacts of using one of these variables might be I don't know Common Language tell them what it means of friends in writing so you can think about it and understand it and then have a conversation with you. Small
Things translate into a language they understand Finding here. We have a really complicated results from a regression model very few people outside of your data team know how to read this or want to read this and think about what's important here in this case supposed to be analysis done on the number of dropped calls from people who left at Elko service for people who stayed inside Cure Insurance customers averaged almost two more dropped calls a month. That's what's important from what looks like a super crazy
output to someone who doesn't know how to beat it. Taking an extra step too much translation means you can start to have a conversation and you are analysis can start decision-makers and how did it go? Wilson what you always be reducing the cognitive load of our audience which producer on the float and do the heavy lifting of interpretation and you're presenting Adidas that they've never seen before you need to help them understand the data that you yourself have been
immersed and think about what they need to understand. And the reason we need to do that is a lot of the visuals we put out in the world or default visualisations that come out of our dashboards and a visualization software don't have that. I default reducing cognitive load if you come across so many visualisations that people think are fun and cool and show their point. But in fact, they're actually very difficult to quickly grasp and understand and they need to be translated in
trying to just be cute overall importance gets lost. Just for fun, and you can share with you at what is the most awful of that truck? I never seen about banana exports that somehow. Hope you have a good night. Cramps right after starting to be overly complicated and sometimes bad sharks are bad for a data analysis. When really what we're trying to do is explain to our audience and you can't be certain that your audience knows how to read a thing or an area chart whisker plot
25 years. The general rule of thumb is when possible and I know this kid is injured for all data sets. But if you can stick to my charts bar charts and Scatter Plots, when you your data, you're going to invite more people into the conversation then if you had chosen a complicated visual, but they don't know how to understand. No one likes to feel dumb. So is he simplify and use common charged with are comfortable with with already reduced their barrier-to-entry
the conversation? What are also invite our audience to experience our expertise? So this is this is a little mold at the Loop bands is. He's written some great are programs. That anyone is a cotton a college sports fanatic really appreciate. Calculate winning percentage for two teams play against each other and basketball and my husband and his buddies have this group chat unless win probability chart does look like it has symmetry is aesthetically pleasing Ryan by the way as an accountant.
Slingbox to create opportunity Luke Benz forgot that his audience might not understand this part. And so this is a perfect opportunity for me to do a little translation by thinking about my audience is Brian group chat. I'm thinking about what was really important for them to know what if you call out what was happening in the game for my audience if you miss me pants with to make it clear and this was a crazy game that Carolina lost twice. I've basically by I having last-minute buckets to do my duke super stressful game, but now
we have a visual speaks to me what I needed to know where was much harder when we are looking at that win probability chart, right? It already just looks too technical. And to finally be considering your audience in considering your message considering your translation. There are many many different ways that you can go with helping your audience understand that they need to move forward and make better decisions and be moved by data. And so I can actually improve your
good subject matter. What I did is I put together a very one-sided be a collection of what it is like to live with Ryan. And totally one-sided not very scientific. I'm plotted the habits that he doesn't turn his bike living with him. How often he doesn't look at the bubble chart here and put the level of enjoyment. when I think about translating this Vita to someone I have some options. So I just met all of you and I mean it certainly tell you. Well, there are
a couple of hours. I was going to show you how great he is instead. We'll look at the second version where it rains my audience if I need to translate for him what he should take out of this data. I'm going to have him focus on this bottom left quadrant things that drives me crazy that he does and I'm going to make a move. I'm going to translate some of my language into his so before it was minimal effort maximal effort. Is it easy? Is it hard and then I had it before on the scale of enjoyable too. Irritating quite frankly. I'm
pretty sure that's not fair if he thought that might be difficult sometimes. Remind me to change the language. I used to understand the level of seriousness of what I'm talking about here. So he's going to go with grounds for divorce free Ronnie's changes from your jargon abbreviation ways that you can consider bringing Outsiders into your data so that they can talk about it. This has the time has flown by so as we think about translating for all audiences ever
went to do this with a little bit of practice making turkey focus on your audience. Ask the right questions before you dive right into the data because of the help you stay out of radicals making sure you can write level of translation to help someone come to a decision around and the impact of what you're sharing and most importantly going in with the mindset of embracing the role of being a translator. It's just a little extra work, but it opens up so many more doors for you. My time is just about up, please don't be a stranger. Feel free to connect with me on
LinkedIn or reach out to me. I'll let you know that dress is there and I have just a couple of minutes. I believe for Q&A questions people have about being a translator before it looks like there's a Kino's coming on after this.
Buy this talk
Buy this video
With ConferenceCast.tv, you get access to our library of the world's best conference talks.