[music]

0:00:05.8 Announcer: Welcome to the Analytics Power Hour. Analytics topics covered conversationally and sometimes with explicit language.

0:00:15.2 Tim Wilson: Hi everyone. Welcome to the Analytics Power Hour. This is episode 227 and I’m Tim Wilson. Now stick with me for a minute, 227, the very show number for this episode was also the name of a pretty successful sitcom back in the day. It ran from 1985 to 1990 on NBC, and it starred Marla Gibbs. Now, if you don’t remember who Marla Gibbs is, and she’s still with us, she’s turned 92 earlier this summer. Well, she really exploded on TV and her kind of regular scene stealing during her supporting role as Florence in The Jeffersons, which ran on CBS from 1975 to 1985. Why am I telling you this? Well, because I’m of an age where I actually remember the show, and I think it’s important that you understand it. Just trust me. So I’m joined for this episode by two co-hosts, one of them, Julie Hoyer. Julie, how are you? Do you remember, have any memories of 227?

0:01:14.2 Julie Hoyer: Unfortunately, I do not. I’ve never heard of that show ever, actually.

0:01:23.6 TW: Well, that’s a good setup. And Val, Valerie Kroll, how are you? And do you have any memories of 227?

0:01:28.9 Val Kroll: I am well. And I unfortunately do not have any memories either. The only thing that resonated there, I was like, I think I’ve seen the Jeffersons, but… [laughter]

0:01:37.8 TW: Okay. At least you’ve heard of the Jeffersons, right?

0:01:41.4 VK: Some familiarity.

0:01:41.5 TW: Julie, have you heard of The Jeffersons?

0:01:43.6 JH: No. No.

0:01:45.6 TW: Oh my goodness. Okay. [laughter] And Michael Helbing will listen to this and think, yes, great, Tim’s now making pop references. They’re only, you know, 30 years old, so that’ll be pretty on brand. I do think another important fact about 227, the sitcom not the show, is that it was actually based on a play written back in 1978 by Christine Houston. But anyway, this rambling backstory, me trying to explain something that was completely foreign to my co-hosts and probably to any listeners as well, was kind of my attempt to simulate the emotional state that our stakeholders sometimes feel when they’re looking for just a simple answer. And we have to draw them to some nuance of the data or its interpretation in this case, Okay, you really didn’t need to know anything about the sitcom 227, but it still seemed like a good setup to me, so I ran with it.

0:02:34.5 TW: But what do we need to think about when we are trying to give a non-technical audience or a non-technical stakeholder enough of an understanding of some complex concept or the nuance of the data or something so that they can actually make an informed decision, they’re not running with scissors, as Julie had made a comment before the show, that’s the topic of this episode, and we wanted to bring in someone who has thought about this quite a bit. Dr. Nicholas Cifuentes-Goodbody is the Chief Data Scientist at WorldQuant University, where he created and manages the Applied Data Science lab, which is the free course that helps 1000s of students kind of learn actually complex concepts through project-based learning. He’s created and taught all sorts of material into a wide range of audience types across many, many different learning environments. And today he is a guest on the Analytics Power Hour. Nicholas, welcome to the show.

0:03:30.3 Dr. Nicholas Cifuentes-Goodbody: Well, hello. Hello. Hello. Thank you so much for having me, guys. I’m really excited to be here.

0:03:35.9 TW: Great. I’m really looking forward to this. Before we dive right into the topic, and this is not something we would normally do, but the premise and kind of the business model of WorldQuant University, kind of what it is and how it works, can you give kind of the quick background on that? ’cause it’s, it’s kind of intriguing.

0:03:54.6 DC: Absolutely. So WorldQuant University is a nonprofit venture. It’s founded by Igor Tulchinsky, who is one of the pioneers of quantitative trading. So I think you recently had a guest who was into financial engineering, right? So those are some of the techniques that he sort of pioneered. And the idea with WorldQuant University is that there are talented and hardworking people all over the world, but unfortunately, opportunity is not always there where they are. And so the question is how can we use technology to open up avenues of opportunity to those talented people, and specifically how do we do that in financial engineering? So we are an online accredited university and we have a master’s degree in financial engineering, and it’s absolutely free and open to anyone. We just have an entrance test. And then we also have the Applied Data Science Lab, and that’s the certificate program that I run that’s a shorter self-paced offering where we go from basically, you know a little bit of Python, you know just enough to be dangerous to let’s build our own API to deploy a model. So that’s what WorldQuant University is. We bring opportunity where it’s not, and…

[laughter]

0:05:14.4 DC: We do that in the data sciences.

0:05:17.0 TW: And it’s, I mean, who’s funding it or how is it, like what’s, how’s that work?

0:05:20.6 DC: Yeah, so, right. So it, so it’s bankrolled by Igor Tulchinsky, right? Who is yeah… And he made his… I mean, quantitative training is sort of where he came from. So this is kind of one of the many nonprofit ventures that he has.

0:05:38.0 TW: Okay. Interesting.

0:05:38.9 DC: So we have our funding and that way when you show up, you don’t need to write us any cheques. You just show up, you take your entrance exam and then we’re just here to help you learn.

0:05:47.8 JH: It’s really cool.

0:05:51.7 DC: Yeah, absolutely. You guys should sign up.

[laughter]

0:05:53.9 JH: I’m tempted.

[laughter]

0:05:54.0 VK: I could use some of those classes.

[laughter]

0:05:56.5 TW: Okay. Well that’s, that’s, that’s fascinating. I’m sure like a number of people are now like, what? I’m gonna go check that out ’cause that’s…

[overlapping conversation]

0:06:03.6 DC: Please, please, you can go to, yeah, no wqu.edu. You can see more about it there. And I really do encourage everyone to come. Currently in the data science program, we have about 7000 students. They’re mostly in Sub-Saharan Africa and Southeast Asia, but we have a ton of students from the US where you guys are situated right now. We even have a few in Australia as well. So we have people really from all over the world. And we would love to have you join.

0:06:30.6 JH: We’ll have to tell Mo’s team too then.

0:06:33.1 TW: Yeah. Mo will listen to this episode by Gulley.

[laughter]

0:06:42.5 TW: So in the actual topic, so it does feel like a lot of what you’re doing, I mean, directly kind of the day job is teaching complex concepts to people who want to learn the complex concepts, but then there’s kind of teaching the right level of complex concepts to people who just need to know something that’s very narrow, and figuring out the depth and the width. Do you have kind of framings that span both of those or ways of approaching that that you…

0:07:13.8 DC: Yeah.

0:07:14.4 TW: At a high level kind of think about?

0:07:18.1 DC: That’s a great question. So we can think… Oftentimes when we think about a classroom, we have students come in the door, you as the teacher shut the door and all the students are there and they need to be there because they’re gonna get a grade and dagnabbit. They need to get their money’s worth out of their education, right? And so it’s very easy to teach students in that context, but all of a sudden when there’s not a door shut behind you keeping you there, when there are lots of different things on the internet pulling your attention, it’s not quite as easy to make someone sit down and drill a con… Like a complex concept into their brains. And so for that reason, the other thing is we have a very wide audience, right? Some people are native English speakers, others speak English along with three, four or five other languages.

0:08:08.2 DC: Other people are just learning English or they use it as their transactional language. And so for us it’s really all about using clear and simple language to explore concepts that are complex. And often what that means is starting small and spiraling up to the more com… The sort of deeper complexities of the issue. And I think this is a… We talked a little bit about this before we started recording. It’s really all about context. It’s giving the learner the context that they need to see why this particular concept is important and they should stick with it, even if right off the bat it doesn’t quite seem clear to them.

0:08:50.0 JH: Would you say that… Because the part that you just mentioned really piques my interest because I’ve always said that I learned best when whatever I needed to learn informational wise was applicable to like a problem I was trying to solve. So do you guys kind of use that same technique and find that it really helps, like you were saying, kind of draw people in and make it worth their time to stick around and kind of get through the nitty gritty part of it?

0:09:14.5 DC: Yeah, I think so. First of all, Julie, let me say that. It’s not just you, everyone learns better when there is an immediate problem to which they need to apply some sort of new tool, right? The best way to learn how to cook is working in a kitchen. And for many years I was a Spanish teacher. The best way to learn Spanish is in a cafe when you’re thirsty and you really need a coffee…

[laughter]

0:09:37.0 DC: And so that’s exactly what we do in our programs, right? From the very… So what we do is we take students through eight end-to-end data science projects. The first is exploring the housing market in Mexico. And then the last is predicting volatility on the bomb based stock exchange in India. And everything we do starts with, here is the problem statement, this is what we need to solve and this is what we need to do to solve it.

0:10:05.5 DC: And so oftentimes, you’ll go to a textbook and everything will be organized by topic. Like, here are the metrics, here’s the R-squared, here’s the, R-O-C-A-U-C, here’s the mean absolute error, here’s the mean squared error, right? And just marching through those, you’re pretty much gonna have your eyes glazed over by the time you get to section two. And the way that we do it is we do it more sequentially, as you would when you’re a working data scientist, right? First you need to connect to that database. What type of database is it? You need to explore that database. You need to extract that data. You need to think about, hey, this data is messy. I need to clean it in order to get it into my model. And it’s only once you’ve built that model and iterated through it that, hey, let’s look at this one particular metric and let’s see how to use it in this particular case.

0:10:52.2 DC: And then we might use it again and again and again. And then it’s really only at the third or fourth time that we might say, hey, remember that metric we’ve been talking about? Let’s see how it compares to this other metric. And so, like you say, project-based learning is a really effective way to take new information and have a reason to put it into your brain. Your brain is always looking for reasons not to learn things, right?

[laughter]

0:11:15.0 DC: And so this is a really good way to give your brain a reason to hold onto that information. And that’s why we use project-based learning in the data science lab.

0:11:24.2 JH: I love that. And this is actually one of the things I was really interested in exploring too. So I have one layer deeper on exactly what you guys were just discussing with the project-based learning. So a lot of our listeners work as an analyst in an in-house setting or maybe some of their at an agency, and they’re explaining some of these concepts to their stakeholders. So less classroom, more the application with those stakeholders I always struggle with, which is the best approach to educate what’s necessary within the context of an analysis or a project that we’re delivering. Or is it better to do education separate with hypotheticals or analogies and then like, let that come together on the project. And I was trying to think through what are some of the pros and cons? And in the situations where we’ve tried to do education alongside a project, a lot of the stakeholders have a vested interest in an outcome.

0:12:19.0 JH: Like they really want that variation to win on an AB test as an example in the simplest context. And so when you’re trying to explain, well, what are all the, the background that you need to know about the power analysis that, that we use, or what is an MDE? Like, they’re like, yeah, yeah, yeah, but get to the answer, get to the good stuff, [laughter] or they’re like not wanting to hear the answer that we give. But when you do education separate from that example, it’s kind of like what you guys were just discussing that it’s like it’s hard to get invested. And so I wonder like how much is kind of washes over them and and it’s not really retained by the time I really need them to recall.

0:12:52.2 TW: Or they skip it, right?

0:12:54.4 JH: Yeah. Well, [laughter] yeah, that too.

0:12:56.2 TW: Or they flat out get to skip it, right?

0:12:58.4 JH: Yeah.

0:13:00.5 TW: Because it’s like, it’s not like sure it’s required. Like yeah, is my manager gonna come tell me that I had to go to that? So…

0:13:03.6 JH: Yeah. So how do you give guidance to your students? I’m curious or on, and when they’re applying some of these skills in the real world, like what’s the best way for them to impart some of this knowledge and what, what have you seen success with?

0:13:15.2 DC: Yeah. Well I have to admit, we’re all data scientists here. We’re all data people, right? And so I do have a bias, and my bias is always towards context and always towards the context of the project. The business leaders are busy. They have lots of meetings and lots of PowerPoints that they need to learn about. And if there is… Let’s do a little side route around, to talk a little bit about this metric or how we do a power analysis within our, AB testing. I might, if I were the CEO say, oh good, this is the time where I can turn off and think about something else and just take a little break from my work.

[laughter]

0:13:51.2 DC: And so I think really if you’re gonna be touching those topics, it’s really in the context of the story that you’re telling your CEO, right?

0:14:00.0 DC: And you really need to think about what are their interests and what are their motives. An executive is interested in two things, time and resources and how they’re gonna allocate those things based on the data that they’ve been given. And it’s your job to give them that data. And so if you think about that is their goal, your job is not to say this is what a power analysis is, but you might say, we did certain techniques that allow us to say with confidence that we… That this route worked better than this route. Or that there was no causal link between the new email campaign we did and user retention. So I have to admit, I have a little bit of bias. I’m always, let’s talk about the project, but I would really love to hear from you guys where are some cases where doing a little bit of side education would work? Maybe it’s not working with the CEO, maybe it’s working with the data team, right? Maybe there’s a… This is a concept that they’re gonna be using over and over again. And so building that up with them is a really successful way to do that. So I’d love to hear about your experiences too.

[music]

0:15:04.9 TW: Alright, it’s time to step away from the show for a quick word about Piwik PRO. Tim, tell us about it.

0:15:13.5 TW: Well, Piwik PRO is easy to implement, easy to use and reminiscent of Google’s universal analytics in a lot of ways. I love that it’s got basic data views for less technical users, but it keeps advanced features like segmentation, custom reporting and calculated metrics for power users. We’re running Piwik PRO’s free plan on the podcast website, but they also have a paid plan that adds scale and some additional features.

0:15:34.2 TW: That’s right. So head over to Piwik PRO and check them out for yourself. Get started with their free plan. That’s Piwik PRO. Alright, let’s get back to the show.

[music]

0:15:45.2 TW: I mean, I’ll throw in ’cause it’s become one of my go-to ones. And, and it’s funny, we’re kind of… We’ve inherently wound up on this kind of string of communication themed episodes. And I think about like when, and this was back when I actually worked with Julie and Val, but I was working with, with Julie on one where we were going to use Bayesian structural time series. And it came down to a visualisation of saying, we’re gonna tell the business person just visually show them the idea. Not by any means diving into Bayesian anything, but just making sure the visual said, here’s the historical data. And that data can have a model and a model has some uncertainty to it. So putting some error bar or kind of confidence bands around it and doing a quick little buildup and like, I can do this in 60 or 90 seconds enough for you to have the intuition.

0:16:42.8 TW: Which to me, the big, big thing was what they want is they want a hard number, they want the answer. And that’s like the bane of the analysts or the data science’s existence is this expectation that there is a definitive truth. And the number of times I now drop, confidence bands, prediction intervals, whatever, subtly ’cause I feel like that’s kind of intuitive and the education is being like, this isn’t perfect. I can’t give you the perfect answer or even the results of an AB test saying, I will show you that there was a 4.7% lift, but by Gulley, I’m gonna put error bars on both of those so that you understand that it’s not as squishy as you would like it to be. I don’t know, as you were talking, that’s, that was again, kind of in some of our prep there was a little bit of like, ah, like trying to explain it to them. We can’t go back to considering you have a fair coin and you flip it 500 times.

0:17:40.8 DC: No, never. No, no, no.

0:17:42.4 TW: That’s not gonna do it.

0:17:43.8 DC: Don’t even talk to your students about coins.

[laughter]

0:17:47.2 DC: There should be all coins should be absolutely banned from statistical education.

[laughter]

0:17:53.5 JH: You heard it here first.

[laughter]

0:17:54.4 DC: I heard here. But yeah, I’m unilaterally declaring a death to coins.

[laughter]

0:17:58.5 DC: Death to coinage.

[laughter]

0:18:00.0 DC: But Tim, that is a wonderful way of thinking about how to engage with your audience where they’re at, right? Maybe it’s not that you need to talk to them about types of statistical certainty or uncertainty, but rather maybe it’s, you need to show them visually, right? That you need to represent that squishiness like you say. And doing that through one visualisation is a much better way to do it than taking them through a little story, right? Sometimes a picture says a 1000 words like you say. Once you as a data scientist give it a little framing, of course.

0:18:32.8 JH: One thing too that I thought helped with that specific example that Tim just brought up was, I was just thinking back to like the timing of the education piece. And it is kind of nice when you get into a cadence with your stakeholders where they’re coming to you with a problem and you get to talk them through a little bit of how you’re thinking about approaching it. Because I think that’s actually the point, Tim, where we usually were able to bring up those slides and say, hey, this is a technique in a way we believe we can come at this problem to help you answer it. And this is kind of like your assumptions going in and we feel like you problem, like fits those well and this would be a good approach for us to help you get your answer. And so it was like, it was nice to do like a first touch base where they’re interested in like you trying to help them. And I felt like they were more open to receiving that rather than maybe getting it through an email showing up for the readout.

0:19:19.0 JH: When they really just wanna get to the number. Trying to do education on top of them, just wanting their results that helps a lot.

0:19:26.1 TW: It’s an opportunity to just say, This is what you’re not gonna get, ’cause if you’re off for two weeks crunching on something and they think you’re gonna come back with this golden… And you’re like, No, I knew we were never gonna do that. But in the absence of any education up front, that can be their expectation, and they’ve probably told 10 other people that, Oh, the data science is doing magic stuff and it’s gonna… And they’ll commit to what’s gonna be included in the results that aren’t. Whereas if there’s that, it’s the… It still feels weird ’cause they don’t really wanna know. It’s the we’re not trying to turn them into analysts or turn them into data scientists, but that’s like this just crazy. And maybe, as you said, it’s the context of, What do you need to know? How can I keep it brief? And are you going to… I’m really, I always think of it is I want them to have the intuition, I want them to understand what’s happening, and hopefully they’re interested in it, they’re like, “Oh, that’s kind of cool,” that if they can get that like, “Oh, well, that’s kind of slick, there’s a lot more details, but that’s kind of slick, I’m more likely to remember it and then it’s repetition, ’cause then the next time when you’re reading it out or going over it, it’s not the first time they’re not coming in cold, of course, I got the chair. To take an easy example. [chuckle]

0:20:52.7 DC: Well, if you set things up right, all your examples will be easy, right? That’s when you know you’re doing your good communicating.

0:21:00.1 DC: Julie, you and Tim, I think both bring up a really good point, which is often people come to you with questions and you are very inclined to just give them an answer… I’m sorry to snap when we’re recording a podcast, but just to give them a snap and answer, right?

0:21:13.1 TW: You’re just showing off the quality of your microphone.

0:21:16.6 DC: Yeah wow. It’s crystal clear. Crystal clear. But really what you should be doing is asking a lot more questions and getting a lot more context, and then using that to set expectations. So one thing that I always tell my teachers is, before you give an answer, make sure you’re asking a question first, and so that’s a really good example of that.

0:21:34.7 VK: I remember when I was really maturing as an analyst or I’d like to think I was I am gonna say. And I was like, I’m realizing now that my analysis is only as good as the questions that are asked and how I can refine that premise and what I’m really exploring or validating. And so I was like, I know what I’m gonna do. I’m gonna do a brown bag or I am gonna do a meeting, it started as a meeting, 90 people I was gonna invite to our top floor, and I’m gonna teach them the title of the meeting was developing a good hypothesis, and I wasn’t getting a lot of accepts to this meeting, so I went to my boss, I was like, “Listen, can I spend 300 bucks on pizzas for a lunch up coming?” And he’s like, “What? For what is this for? And I was like, “Well, I’m doing this meeting, and this is what I’m hoping to do, whatever.

0:22:19.7 VK: He’s like, listen, I hate to break it to you Val but there is no amount of pizza, to get people to show up to developing better hypotheses. He’s like, We have to find a better way to get at this. But I was like, oh, okay, but that was just kind of learning how to meet someone where they’re at. It was a good lesson for me to learn.

0:22:39.4 DC: So what did you guys come up with? I’m curious, what’s the resolution to that story.

0:22:43.9 VK: So it’s actually very much what we were just talking about is outside of the context of the problems that my different stake holders were trying to solve… They weren’t really interested. It’s like when someone tells me about a dream that they’re not in, and you’re like, Okay, boring can’t wait to talk about my weekend. But we decided to take that in the slides that we created for that training, where we were talking about the template, like where it’s important to state what your expectations are and what metrics you think will move and what’s your rationale, and that became like a little bit of a preamble before we got into project-based sessions. And so that was where we were able to talk about something in the context that really meant something more to our stakeholders, so no brown bag lunch on that format. [laughter]

0:23:26.2 DC: You didn’t even need pizza for that one, right?

0:23:28.1 VK: Exactly. It was free.

0:23:31.0 TW: But Val, when you’re going into… If it’s somebody who’s kinda knew who’s coming in with the, we wanna run an AB test, like that starting point, we need to be doing AB testing, and I feel like lots of people have the basic basic intuition of AB testing before they’ve gotten the cold shower of, you need enough data and you have decisions you need to make around… These are the levers you can pull and you’re pulling them in through guess work. Do you find that that is an opportunity? If you say, great, do you… One-tailed or two-tailed does it need to be… And you’re not necessarily asking them that, but you can ask them the questions to try to figure out what kind of test and along the way, say, I’m educating you because… Not explicitly, but try to gently educate them that the trade-offs they’re trying to make, ’cause planning a test experiment is all trade-offs and what level of certainty do you need or how and what are you willing to spend on it, and that’s to me a complex, that’s the quintessential complex concept of, these are all the levers you can pull. You can run the test longer, you can simplify what you’re testing, you can change your minimum deductible effect, you can change the confidence, you can change the power.

0:24:54.5 TW: That’s complicated, and they kinda want you just to do it, and where do you meet where you’ve probed them enough to make those judgment calls on your own, but then you have to always do that every success of one versus how often are you saying, Okay, can we just talk about one-tailed versus two-tailed, and it’s one diagram of a normal bell shaped curve. How do you approach that? And anyone…

0:25:29.5 VK: I’ll give you… Juliana, we were actually talking about this.

0:25:31.8 TW: I was like you… I was asking you, but. I was like, That’s unfair I am like cul de sac.

0:25:41.4 VK: I definitely have something to share about how I approach it, but Julian, actually, Julie and I were talking about one-tailed versus two-tailed, and what do you reveal to the client earlier today? I swear, I’m not just like… We were talking about it earlier today. No, we actually were.

[laughter]

0:25:54.7 VK: But it really depends for in our situation, like who our primary stakeholder and client is, Are we supporting a data and analytics team where they have some of this base understanding, even if they’re not super into testing it, or are we working with product or marketing stakeholders and so when it is the product marketing stakeholders who are less aware of some of these concepts, nor do they have some of the baseline things you can kind of call upon to weave some of these things together? It really is just a discussion and keeping it really conversational and saying things like, How would you know that the alternate version that you’re proposing won or is a winner? Or how would you define success? And so then that’s when you start to back into… It’s like playing, Guess who, I can ask a couple of questions in this logic tree and it backs me into some assumptions, and then we can kinda turn that back around in a test plan and when we walk them through that and say, So based on our discussion here is what we’re thinking.

0:26:51.5 VK: They’re like, What? That’s gonna run for eight weeks, I have to launch this campaign in four, so then we’re like, Okay, well, with that constraint, here’s our options, and so sometimes you talk about trade-offs there, but if you can get to a starting place with all of the parameters in an assumptive state, then it’s easier to show how one thing moving at a time can affect other things because it is nuanced to your point, but that’s one thing that’s worked well, is the conversational discussion guide.

0:27:20.6 JH: Well, Nicholas, I wanted to ask if you have any tips or tricks or things to think of when you run into someone, ’cause to Val’s point. I almost feel like we end up talking to people who know enough to be dangerous, like they are decision makers, and they know enough about what we’re talking about here when we’re discussing the one-tailed versus two-tailed they understand enough of the concept, but they may be making a decision-based on incorrect reasoning, and it’s hard to… It’s almost harder, I feel like working sometimes with those clients or stakeholders than if we were just working with marketing stakeholders who are trusting us to actually make those testing, like decisions for them or modeling decisions for them. So I’m curious your thoughts on some of that.

0:28:07.2 DC: Yeah, I think… I mean that’s a great question, and I think what you need to do is approach your clients or stakeholders in a way where you’re giving reverence to those ideas, but not in a way that it’s shutting down conversation, because if everyone thinks, Oh, this is a big important idea. And we all know what it is, it’s fine. Then people start, like you say, running around with scissors, so one way in which I’ve seen this happen a lot, is a data scientist will come in and they’ll start using words like, Oh, well, this is easy… Well, we just do this. We simply do that, and what you are then… What you’re trying to do as the data scientist is you’re saying, let me make this accessible to you, Hey, this is fine, this is easy, this is fun, right? What the other person is actually thinking is they’re thinking, Oh my God, this doesn’t strike me as easy at all, if I don’t understand what’s going on, I must be a total dumb, dumb. And so then what they wanna do is they won’t ask any questions, and what I’ve found is that it’s much more likely that a client will simply leave a meeting not with a dangerous idea of what to do with a mathematical tool, but with no idea, and they’ll go, Yeah, this was fine, the data analyst it and like it was… I don’t know, they thought this, right.

0:29:20.8 DC: And so really… But if you can approach it in a much more way in which you’re thinking about the context, like we can think about how we’ve been developing this conversation in the last little bit chunk of time, right. Thinking about the context of your stakeholder, thinking about where they’re at, making sure you’re asking lots of questions, what does success look like? What are the levers that we can move here, what do we have control over, what we don’t have control over, then you can hear them start talking about the topic and you’ll hear them making assumptions, and so that’s when you jump in with that Tim moment, you go, wait a second, that’s an assumption here that we can’t quite make… Hold on, let’s just think about this as well. And you never do it in a way where you go, you’re wrong, you always go excellent…

0:30:10.3 JH: You will never win friends that way.

0:30:12.3 DC: Let me take what you said and let me just re-arrange just a couple of things and build on it a little bit. Alright, and so that’s where you put your slight little corrective in there, and so that’s what I would say is the best way to avoid anyone using these these Weapons of Math Destruction irresponsibly.

0:30:30.4 TW: Try my…

[overlapping conversation]

0:30:36.7 DC: Absolutely. I never said. I never said.

0:30:39.1 TW: Absolutely.

0:30:39.2 DC: Is making sure that you’re having a two-way conversation with your stakeholders, so that’s what I would say, but I’m curious, you probably spend a lot of times in rooms with clients who think they know what’s happening, or maybe don’t think and don’t wanna tell you, I’m sort of curious. What are some of your experiences?

0:30:56.4 VK: Well, the one thing I will bring up first, experience-wise is when we’ve done a good job of bringing them like they have a problem, we solution it, we bring them along the education just enough, we get them an analysis, we answer their question, and it’s a great outcome. I struggle with… Then they come back thinking, Oh, that worked so well, I got a result I was really happy with. You can just do that again for a problem I give you.

0:31:20.3 DC: Again.

0:31:20.6 VK: And that’s where it gets…

[overlapping conversation]

0:31:21.8 DC: I have my hammer.

0:31:24.8 VK: But I will say that I think just going through the process, even once or twice, or to your point, once you can get them opening up and discussing with you and not feeling like they have to know it all or not ask questions, or they’re starting to realize the more context I give you, this is a good partnership.

0:31:41.8 VK: That education comes a little more naturally. And I definitely have better luck when we’re able to do that kind of natural discussion, I’m listening for red flags, able to do a little more just like course correction rather than kind of pre-defining everything and they feel like, Oh, you’ve got it, and I don’t need to give you anything else and they kinda clam up.

0:32:03.7 DC: Yeah, sure. It seems like the key word, and what you just said was partnership, it’s establishing that partnership with your stakeholders.

0:32:10.3 VK: It’s a good call out. And I have one for you guys, ’cause this is a build on that.

0:32:12.8 DC: Let’s do it, let’s do it.

0:32:14.8 VK: You ready? I remember, and this is one example of this being in a conversation with a client asking about, How do we handle this section testing within the scenario where there’s really low volume, like what are some other ways that we could use data to help inform this decision knowing that we won’t be able to run some of the standard AB tests, and so I started bringing up some different sources of evidence that we could gather, and one of the examples was user testing, and then two weeks later, I joined a meeting and someone else from that person’s team was like, Oh, I heard that we don’t have to do AB test, we can just do user tests or we can do pre-post with user tests.

0:32:53.5 VK: We don’t need to do AB testing and I was like, oop! Hold the phone. I’m not sure how this game of telephone got back to you in this way, but that specifically I’m interested, ’cause Nicholas, what made me think of this as you were saying, getting your stakeholders or partners to kind of talk back with you so you can kind of refine their language, like, what can we do to encourage that more, to make sure that the narrative is preserved as close to the way that we would tell the story if it was a first-hand experience?

0:33:23.6 DC: Boy, that’s a tough one.

0:33:24.0 JH: That’s a good one.

0:33:24.5 DC: That’s tough one. I think oftentimes, we forget that we need to say things 10 times before the message really gets out there, we are the kind of folks where it’s like there’s an equation, it’s right or wrong, or it’s not… We say it once, we’ve got it, but really when it comes to communication, and this gets back to the communication team for this episode, you really need to repeat over and over and over again for something to actually get out there because people have a lot of noise in their lives.

0:33:54.7 DC: And so it’s not gonna land the first time you say something. This is actually big for me in the classroom, I often think if I just say it once the students will get it, right? If I just say it once really, really slowly and really clearly, they’ll get it, and of course that’s not true, because half the time people… Half of the folks were on another Zoom call or they were thinking about something else, and it’s really only after repetition, spaced repetition that… And getting people to recall that information that you can really push it out there.

0:34:24.3 VK: It’s somehow.

0:34:25.5 DC: So sometimes you need to repeat something so many times you’re like, I can’t believe I’m still repeating this, but it’s important.

0:34:34.0 VK: It’s funny how much words matter in math, like you think math is all numbers, but when you’re talking about math and you’re talking to people on how you’re applying it, words matter so much, one word tweet can completely change what you are trying to get across, like the idea or the assumptions or the outcome…

0:34:51.0 DC: Right, and it might be something that’s totally inconsequential to somebody who’s not familiar with the discipline, they might just throw off two different words, and for you, it’s a big deal for them, it’s just a little bit of a flourish with the authors of authors. However you say that.

[laughter]

0:35:08.2 VK: We’re picking up what you putting down.

0:35:09.1 TW: Let’s pour one out for the word for the word confidence. Which gets used in eight, nine different…

0:35:14.3 VK: There’s one more thing I would love to hear you chat a little bit more about Nicolas, you said early on in one of your answers to our questions, you were talking a little bit about motivations of the stakeholders or the people that you’re meeting with, and I love that so much. And it’s something that it’s easy to forget about. Bit it’s so important. So I wonder if you can talk a little about that, especially when the people in your audience have a diverse set of backgrounds, especially culturally, if you’re working at a larger global organization, I’m sure you have some experience there that would be really valuable for our listeners to hear.

0:35:44.9 DC: Yeah, I think this is a great example of making sure you are meeting your audience where they’re at, and so being clear about what everyone’s interests and motivations are, what’s their prior knowledge, what do they not know, what is it that you need them to know and what is it not important that they know the stuff that they don’t need to know. And so I think that the best way to work in that context where you have a diversity of stakeholders in a meeting, maybe they’re both diverse in terms of what their interests are, maybe they’re diverse, both in terms of their cultural backgrounds, the one way to really make sure that you are constantly where they need to be in terms of communication is to check in with them, and so speaking slowly, pausing, asking questions and giving time to breathe, so oftentimes what we do when we are presenting is we say, Okay, any questions? Nope. Alright, let’s move on.

0:36:46.4 DC: So what we say is, if you’re in-person and you’re working… What we say when we’re working in the classroom is, if you’re in-person, you need to wait eight seconds and you really have to count one Mississippi, two Mississippi, three Mississippi. Otherwise, you’ll breeze over it and then… And this is really, really important. People forget this, if you’re doing this online, we’re all presenting much more on Zoom now, right, you need to wait 20 seconds and it will feel like an eternity to you.

0:37:13.9 VK: 20?

0:37:15.2 DC: But you really do, because what happens is there’s a three-second lag, they need to look at your slide, they need to consider what you said, they need to catch up, they need to move their attention from their email back to you. And so taking time, being slow and asking questions is really important, and when it comes to intercultural… Oh yeah, go ahead, Tim, I’m sorry.

0:37:39.9 TW: Well, I guess that’s the other on the waiting ’cause it is uncomfortable and I am definitely uncomfortable with the pregnant pause, but if it’s a group and you ask and you wait for 20 seconds, it seems like guaranteed somebody will be so uncomfortable with it that if they’re also uncomfortable asking a question, they’ll ask the question, they’ll kinda break the dam… In the first question, it’s like nobody wants to ask the first question, but if you wait long enough and that one person who just has whatever psychological issue that they can’t… They have some question, it may not be the best question, but it’s probably a legitimate question. They’ll ask that and then everybody else is like, Okay, well, if he’s gonna ask, I’m gonna ask my question too. I mean that’s… I’m just filing that away is like the… I’m gonna sit and wait.

0:38:30.0 TW: And I am gonna be dying. They’re gonna think that I’ve locked up, but if I get the first question, then I’ll get the second question and then maybe we get back to have… We have a little discussion and repetition and the concept gets to sink in a little bit more.

0:38:45.3 DC: Yeah, another thing you might do…

0:38:46.9 TW: Yeah.

0:38:47.2 DC: Is sometimes the problem on the asking a question is asking it to the person who’s presenting, people feel like they feel they’re intimidated or they don’t wanna interrupt or be rude, so what you might do is break everyone up into groups of three, and you might say… Or groups of two and say, Alright, I want you to as a group, what is one question that you had that you weren’t sure about at this point. And so on they one hand, if you’re really lucky, everyone will answer their questions among themselves, and if you’re extra special, lucky, you’ll see that they all have questions in common, which means up there was something that I missed that I really need to address here, and so it’s really, really important sometimes you just wanna keep moving and just dive through that presentation, just get it done, get the slides done, but pausing even when it’s painful. Pausing when it’s painful. That could be the title of this episode really.

[laughter]

0:39:41.9 DC: Is a really helpful way to check in with your audience and make sure that you are at the level that they need you to be at as a presenter.

0:39:49.2 VK: That is such gold advice about breaking the small group, asked questions. ‘Cause that also check a box of the validation of explaining to someone else your understanding it really even solidifies it more with you so that’s… I can not wait to try that one. So that’s a good one.

0:40:06.7 DC: Great, well, let me know, I wanna hear about it.

0:40:09.6 VK: And I will say that in 2019 or 2020, I did have a goal for myself to be as comfortable with the pregnant pause as Merritt Aho, I don’t think I got there, but he was the King, the King of the pregnant pause so shout out to Merritt, if you’re listening to this one.

0:40:26.1 TW: Oh my, I wanna use the word earlier, and it slowly marinated in my brain, you said reverence, and I may take this in a different direction, but I’d love to get everyone’s thoughts. There are things that I have learned as I’ve grown as an analyst that just blew my mind, like my go-to is time serious decomposition. I will go to my grave thinking that that is just wild.

0:40:55.7 DC: What is it about time series that’s so compelling? I know that… I know we’re talking about talking to non-technical people or of a second, we could just say, how awesome is time series data, right? Is it just the coolest thing in the world? Yeah.

0:41:08.5 TW: Well, but it’s also tough ’cause people wanna take it and they have no concept of stationarity and they wanna do just a scatter plot.

0:41:18.2 TW: That’s one of those… Another one that I’ve gone through like saying, This is why this is problematic. And first differences, this is another one that I just go to my grave, I’ll be on my death bed and people will be like any final words and I’m like just talk about first differences, you said makes data more stationary. And so…

0:41:33.0 VK: He’s not joking Nicholas, he’s not joking.

0:41:36.6 TW: I’m not… Yeah.

[laughter]

0:41:39.1 DC: I don’t doubt. He seems very serious.

0:41:40.6 TW: But I…

[laughter]

0:41:41.4 DC: Your listeners can’t see, but he has a very serious face on right now.

0:41:45.8 TW: But there are things that communicating some of these concepts, they’re so cool to me as an analyst, and it’s not one where I struggle ’cause I’m like, I think they’re really cool ’cause this is the field that I’ve chosen and been drawn to, and they’re cool. Now, I’m also showing enthusiasm and energy, which is probably gonna make me be more receptive, but does that make sense? It feels like a challenge. I have reverence for some of these techniques in a way like, Oh, we should come a lot, you should… You wanna get some intuition about this because it’s so cool, and reading to the right level of what that intuition is or what… Using that excitement to the benefit and not like, Oh crap, the nerve is just gonna drown on ’cause he’s geeked out about something that I don’t really need to know, does that make sense as to where to fit between those?

0:42:41.6 DC: Well, maybe if we thought about it in the context of other art forms, for example, in a kitchen, there are things that we see when a chef is preparing a meal, maybe we don’t wanna know about every single little French technique they’re using, but we can understand the excitement and the artifice and the beauty and deliciousness of their food, or the same thing with a painter or an actor. We may not know all the techniques that they are using, but their excitement is helping us… Helping bring us into a world, that we don’t really know anything about, but their excitement is giving us an appreciation of it. Does that make sense?

0:43:24.8 TW: I like it. I wanna think… Yeah I think, I can think about it that way and might keep myself out of trouble potentially.

0:43:34.1 DC: Often, I think about this is you try to explain these concepts to someone as if you were at a cocktail party and you were talking with somebody who had no relationship to the things that you do. So how would you do it in a way that’s quick and exciting for them, but within the bounds of, We can’t talk for more than a few minutes, I don’t wanna be… I don’t wanna corner you here in this cocktail.

0:43:57.1 VK: You gotta gauge their interest.

0:43:58.1 DC: Yeah.

0:43:58.6 VK: Before you go any further, right?

0:44:02.0 TW: Awesome.

0:44:02.6 VK: They’re willing to spend five minutes or 10 minutes?

0:44:03.0 DC: Yeah, this is like when people talk to me, when someone’s like, Tell me about natural language processing, I’m like, Can I talk to you about embedding of words, can I talk about vector representations of words?

0:44:12.7 TW: This is a… Oh my God.

0:44:12.9 DC: And they’re like, No, sure absolutely not, you can’t.

0:44:17.8 TW: We had a Webinar next Wednesday. Where the embeddings API, the speaker went over it, and that’s another one.

0:44:25.0 DC: Yeah.

0:44:25.2 TW: Completely blew my mind. I’m like, This is so cool. So unfortunately, my wife and daughter caught a little bit of the brunt of that over the last 48 hours, and I’m like, literally, you throw anything at it and it gives you a vector that represents where it is in the model, so I stopped.

0:44:41.0 DC: Yeah.

0:44:42.0 TW: Okay. Clearly I have not learned this thing, I don’t understand it, that’s still… But it was wild.

0:44:48.0 DC: Maybe what you need to do is you need to think at this point, would my wife and child be rolling their eyes, is that… And that’s maybe an indication to roll in to really your section a little bit.

0:44:55.8 TW: I’m 30 years in. I’m not sure that the pattern of behaviour are gonna change in my relationship. Well I…

0:45:06.1 VK: I was gonna…

0:45:08.7 TW: Whoa, oh you wanna… You were gonna play the more role and have one more thing as I’m trying to have you rap? Okay.

0:45:16.1 VK: Yes I am. Because we’ve heard keep it like a cocktail hour entry level, come with enthusiasm for your art to get them to wanna buy in and appreciate it, but we also talked about… We use a lot of slide where or Tim did the example of the visual for that Bayesian time series model he was talking about, and so I was just wondering if you had any other do’s and don’ts or best practices for the visualizations or any other key takeaways for best ways to communicate these things, ’cause I rattled off the first couple.

0:45:49.3 DC: Yeah, well, I think those are all great. And they must be great, I said them.

[laughter]

0:45:55.8 DC: I think we’ve been talking a lot about language here, and keeping our language accessible is one thing, think about intercultural competence with all of your stakeholders, you don’t wanna be in a meeting with a bunch of stakeholders from India throwing baseball metaphors at them. So think a little bit about the cultural context that you’re operating in as well, and then you also mentioned visualizations, we talk about accessibility in our language, we don’t always talk about accessibility in our visualizations, so when you’re doing your visualization remember, there is a good portion of your students, there’s probably somebody in your audience is probably a man and probably has some sort of color blindness, and so make sure that you are working with colors that are easily contrastable for many different vision spectrums, so that usually means blue and orange, and even think, how would this visualization look if it was in black and white, would somebody still be able to understand what’s going on here? So think about the accessibility not only of your language, but also of your visualizations.

0:47:00.0 VK: That’s a great one.

0:47:00.5 TW: I love it. Think about Deuternopia and then don’t talk about unobserved heterogeneity. Sorry, I was trying to come out with the…

0:47:10.6 DC: Yap. That’s it. Thank you, I couldn’t have said it better myself.

[overlapping conversation]

0:47:12.5 TW: Most of my time obscure references…

0:47:16.2 DC: This is where your wife and your… This is where the eye rolling begins.

[laughter]

0:47:24.2 TW: Yeah. It’s… Again. Well, I feel like we could talk for another hour, but unfortunately, time is ticking on, so we’re gonna have to head towards the last part of the show, this has been a great discussion, but before we go, we always like to go around the horn, go around the screens and have everyone share a last call, a tip and idea, an article, a book, a movie that they have found interesting related to this topic or not. And Nicholas, you’re our guest, would you like to start us off?

0:47:58.3 DC: It would be an absolute honor. So recently, I came through… I had a medical issue where I was unable to read or write or look at screens, and so that meant for several weeks, I was listening to every podcast, every single book on tape in the world, and there is a book that I listened to that, I really wanna share. It really is the best thing that I’ve read all of this year, and it’s called Mathematics for Human Flourishing by Francis Su, and you can find PDFs, it’s easily accessible, both in tape and as a PDF, but what it is is Francis Su is a mathematics teacher and he talks about how Math teaches many different values and there are values that you wouldn’t normally associate if you’re thinking very traditionally about math, so it talks about exploring, play, beauty, truth, power, injustice, and it really is a call to action to see math in a more open expansive way, but also to teach in a more inclusive way, and so it springs from this dialogue, this correspondence he has with somebody who’s studying math in prison, and it throughout that conversation how these themes emerge, and so I just can’t recommend it enough for you guys, I know I don’t need to convince you that math is cool, but this is the book that you could give to everyone, and they would say Not only is math cool, but its values are my values.

0:49:26.5 DC: And so for me, I think it was a really always a good… I sometimes find math a little bit intimidating, it wasn’t my original background, so this is an example of where I was really like, Yes, I wanna be part of this club, so it’s an inclusive and a wonderful book, and I can’t recommend it enough.

0:49:42.5 TW: Okay, so Julie is the person on this call who has a master’s in math. Did you have any courses where you felt like you were learning about values?

0:49:53.4 JH: No, but I agree, I love my math background because I felt like it gave me a lot of applicable skills and I… Yeah the more I’ve gone into my career, the more I appreciate my Math background, and I love math forever and always, so I wanna read that book.

0:50:09.8 TW: Awesome, and would you like to share a last call as well then, since we’re…

0:50:13.4 VK: Sure, so I feel like I’ve been a little bit on my last calls on a AI kick, which is not like… I don’t know very me, I didn’t really see that for myself, but I happened to listen to a podcast, which I’ve actually done, as a last call that the whole series, Allergies, one of her latest episodes was actually about neuro technology, and so it was about AI plus brain technology, and one I just thought it was so interesting because of the combination of talking about AI specifically with data coming from our brains, and she gives this overall view of AI outside of when we talk about it in industry, and it very much got me thinking and I’ll probably go back and re-listen to it.

0:50:53.8 JH: Because the guests on there, her name is Nita Farahany, and I guess she actually wrote a book that came out earlier this year that I would love to read. I don’t have the title written down, but she mentions it in the podcast, and it sounds amazing, but she talks a lot about what is conscious freedom, the idea of experiential learning for AI, is there some knowledge that it can’t gain because it’s not a human experiencing certain things, she also talks about the idea of, it’s very much a gray space, how can it be a tool for education, how could it actually hinder people’s education, keeping them from learning how to be critical thinkers, but anyways, it was really cool and they actually said the craziest thing was they are starting to embed sensors and things like air pod, not air pods like Apple specifically, but earphone’s headphones that may have sensors that can pick up data from your brain.

0:51:44.7 JH: And I was mind-blown, so just like a whole can of worms…

0:51:48.0 VK: Whoa.

0:51:48.1 JH: That got opened up with this podcast and I suggest it for anyone interested.

0:51:53.9 TW: Wow. Sounds awesome. Val, do you have a last call?

0:52:00.0 VK: You bet you I do. So I hope I’m not late to the party on this one, but my mind was pretty blown with this Baymard Institute. I hope I’m saying that correctly, but the connective tissue to this episode is Baymard Institute does an excellent job breaking down UX principles and concepts for people with a lot of great visual examples, so they have a lot of free research and benchmarking reports, and so a lot of times, especially if you’re in CX and you’re approaching that from an analytical point of view, you’re gonna be asked to make some recommendations or to give some input on some alternate variations of the customer journey, and so this is a way for you to look at across 250 almost e-commerce sites at all different aspects of the journey, from faceted search to selecting dates in a calendar, and it benchmarks them in an interactive visualization, and it shows you what was average and what’s outside average, and you can click straight to sites from Nike to travel and hospitality sites, but it’s been an incredible resource for thinking about some competitors or out of category heuristics for some of our analysis that we’ve been doing and I… You talk about Wikipedia holes, oh my God, I’m like 34 tabs deep, and I’m like, Oh my gosh.

[laughter]

0:53:14.9 VK: So I’ve got to stop, but it’s been incredible for education and myself and for myself, but it’s also for my clients, so it’s been awesome. Baymard Institute.

0:53:24.4 TW: Wow.

0:53:24.8 VK: And how about you, Tim, do you have a last call today?

0:53:26.9 TW: Well, after those, I don’t know that I wanna even try ’cause those all seem pretty amazing. Wow, I’ll deal it the most Tim thing, and I’ll do two. One, I was not intending to do, but because Nicholas brought up, we were at the very end talking about the data visualization piece, it’s adjacent to that, But Kate Strachnyi, she’s the… She’s dedicated, is the… Her brand on LinkedIn, but she is on an episode of the Women in Analytics After Dark Podcast, hosted by Lauren Berk, and they recorded it at the Data Connect Conference this year, but the discussion was around building and growing a personal brand in data, and it was super refreshing ’cause it was the other extreme from the figure out your voice, figure out your brand. It’s a little… She’s hilarious, but she’s also very much a data visualization kind of person as well, so I got to actually watch it being recorded live, but it’s now also available on the feed for the Women in Analytics After Dark.

0:54:37.0 TW: But then the other thing, because I do feel like throw something in on the AI front, and this was a listener recommendation following a last call from several episodes ago, a listener recommendation also a friend of ours, Pauline Gaynesbloom. Pointed me to recast, let’srecast.ai.

0:54:55.0 TW: So not to be confused with the recast media mixed modeling company, but it’s an app and it basically takes articles and turns those into much shorter conversational podcasts that are pretty good, I do get an appreciation of good writing that if there’s something really well written In The Cut, that’s a 13-minute read and it turns it into a six-minute discussion where it hits all the points, it’s just a little wild From A Natural Language Processing converted to voice with some level of intonation and doing summaries. You can also point it to things that you want to turn into a podcast for you, there are some character limits on that as well, but if it’s like, Hey, I saw that post, Ben Stansil’s latest post. In theory, I could throw that at the app and it would say, Oh, I’ll turn that into a conversation and you can listen to it, so just another one that’s wild. So there’s my two for trying to make up for…

0:56:01.4 VK: Fun.

0:56:02.8 TW: Match any of those. Amazing ones.

0:56:06.9 DC: Nice.

0:56:07.2 TW: Alright, well, with that, we are bringing ourselves to a close, this is where Michael Helbling is amazing, and I’m gonna fumble my way through this, but this was a great discussion and I’ve got all sorts of things to think about. Nicholas, thank you so much for coming on and having this discussion and sharing your thoughts.

0:56:30.3 DC: Oh. Thank you, it was a total honor to hang out with you guys.

0:56:32.3 TW: Awesome, so no show would be complete without thanking our producer, Josh Crowhurst, who is gonna take a little blips here and there and clean this all up so that we… Anything awkward that I did will be removed. We always love to hear from our listeners. We’re easy to track down. You can connect with us on Twitter, through LinkedIn, you can hop into the Measure Slack. Nicolas, if somebody wanted to reach out to you, do you have a… I don’t think you’re on the Twitter.

0:57:03.4 DC: I’m not on the Twitter or the X or whatever it is, these days.

0:57:05.6 TW: Oh the Xs yeah.

0:57:07.9 DC: But I am on LinkedIn, and so really, if you just search for Nicholas Cifuentes-Goodbody or if you’re not sure about that, you can just do Nicholas Goodbody. You’ll find me, I’m on LinkedIn, I’ll do occasional posts there and then on Medium about the things that I’m learning and what I’m finding interesting, ’cause I feel like if I’m gonna ask people to learn things, I should also be learning things myself and so yeah, LinkedIn is always the best way to reach me, and I check it regularly.

0:57:34.4 TW: Outstanding. Alright.

0:57:38.0 DC: And then I’m sorry. Oh wait, wait, one more thing, and let’s not forget, the best way to reach would be come to wqu.edu and you can sign up for WorldQuant University’s Applied Data Science lab, or you can do our Master’s in Financial Engineering, whatever floats your boat. And we’d love to see you there.

0:57:56.8 TW: What’s the… What’s involved in the application and qualification? How involved in that is that?

0:58:01.6 DC: Yeah so for the Applied Data Science lab, it’s a 12-question multiple choice test, and you get two chances to do your best, and if you get over… I think the threshold is 66%. You’re a member of the lab. And for the Masters, it’s a little bit more involved. It’s called a quantitative proficiency test, and that’s just a funny way of saying it’s a longer test that you take, it takes about 20 minutes, but it’s just those two entry tests and then you’re… You can hit the ground running.

0:58:28.8 TW: Wow, that’s amazing. That’s my last call.

0:58:32.3 VK: It’s a great resource, I love this.

0:58:35.1 DC: Excellent. Wonderful.

0:58:35.3 TW: Yes. That’s amazing. Alright, well, this has been a lot of fun, super informative, and for all you out there who are trying to learn complex concepts are trying to explain complex concepts, whatever you’re doing for Julie Hoyer. Val Kroll always keep analyzing.

[music]

0:58:58.9 Announcer: Thanks for listening. Let’s keep the conversation going with their comments, suggestions and questions on Twitter at @analyticshour on the web at analyticshour.io, our LinkedIn group and the Measure chat Slack group using for the podcast by Josh Crowhurst.

0:59:14.5 Charles Barkley: Show smart guys who want to fit in, so they made up a term called analytics. Analytics don’t work.

[music]

0:59:23.8 Kamala Harris: I love Ven diagrams, it’s just something about those three circles and the analysis about where there is the intersection right.

[music]

0:59:33.7 TW: Damn it things that I didn’t get recorded. I was like, son of a bitch.

[laughter]

0:59:42.1 TW: We literally just moved to this platform, the one we were on before you hit record and one, it was subtle. Everybody didn’t get the count down so I could sneak it in and two, if somebody started to say something, I’m like, This is gonna be good, but I knew…

0:59:58.0 DC: Well, we can always start over and get it again, but I just for posterity sake…

1:00:02.6 TW: So what wonderful things were you saying.

1:00:03.0 DC: I was just saying how organized and excellent, everything was.

1:00:11.3 JH: It’s very much a gray space, how can it be a tool for education, how that it actually hinder people’s education, keeping them from learning how to be critical thinkers.

1:00:23.5 TW: Wow. I think Julie you have to pick back up somewhere else. It’s amazing, and that was a…

1:00:36.0 JH: We have an Alexa in the corner and my husband decided to play a song, not on the…

[laughter]

1:00:44.0 TW: Keep it in.

[overlapping conversation]

1:00:50.7 TW: What experience? I like that you had some experiential learning.

1:00:54.0 VK: Experiential learning. Yeah she did. Contextual like the context.

1:00:57.9 TW: Rock flag and pause when it’s painful.

[music]

Podcast: Download | Embed

Subscribe: RSS