The meaning of data fluency

- [Instructor] Maybe you've traveled out of the country and you've gone someplace where you don't speak the local language, and that can have its own special rewards, as well as its own challenges. Not too long ago, my family and I spent some time in Japan. We went and stayed in a beautiful traditional home in Kyoto, a little like this one. And our host Koichi really was a delightful man, but he spoke no English and we spoke no Japanese. Now it didn't make communication impossible. It meant we all had to pull out our cell phones and spend a lot of time typing stuff in and translating it for one another. It more or less worked, and I think we got what we needed done, but it was a really slow process and we could never be 100% sure that we were communicating accurately. Then again, we had a wonderful stay in Japan. We all got home safely. So basically it worked out well. So basically our communication was successful in the important ways, but it was probably a lot harder than it needed to be. This is an idea of language fluency that helps set the stage for data fluency. But let me talk a little more about language first. People can speak languages with differing levels of ability. One method for describing these different levels is called the Common European Framework of Reference for Languages, or the CEFR. This system gives three levels of linguistic ability. The first one is basic user, and that means that you are able to do familiar expressions and basic phrases to meet very basic needs. You don't actually have to read out of the book necessarily. You know how to say, where am I going, where's the bathroom, and so on and so forth. It can get you through. It's awkward, it's difficult, but it works. The next level is what they call the independent user, where you're able to have common conversations, as well as discussions in the technical field. So for instance, maybe you're going to give a talk at a conference or meet with clients in another country, you can then conduct the conversation, but it's probably going to be on the stuff that you practiced. And then the third one, according to the CEFR, is the proficient user. This is a person who can communicate spontaneously and with a great degree of nuance on complex topics and basically anything. So that's the native speaker. Now, the CEFR breaks each of these down into further categories, but it gives an idea of the different levels of communication ability or linguistic fluency. And this model can be adapted to work with data too. So for instance, this is purely my own list, but I think it helps frame the conversation about data in fluency. The first level is the ability to work with everyday data. That's where you're simply trying to see common patterns and events, may not even be calculating anything. The next level is data within a profession. This is where you conduct data analysis as part of a larger, but non-data focused responsibility. So for instance, you might be doing marketing analytics. You have to work with data, but that's not your job. Or you are a professional or academic researcher in a topical domain, you have to be able to speak the language, but it's not what your job is about. And then the third one, I'm calling data as a profession. This is for people who work extensively and daily with complicated data as their main professional role. Okay, so those are three different levels. People who maybe are even using data just intuitively, people who use it as a tool, and people for whom that is what they do. Now, I also like to think you can simplify these a little bit more into two word phrases. The first one is everyday data. We're talking about the ability to work with that is data literacy. You can read the data, you know what it means, and you can get what you need out of it. Data within a profession. I like to call that data fluency, which is the name of this course. And part of that's because I'm going to be using elements of data literacy and data fluency throughout this entire course. And the third step, data as a profession, maybe you call that one data science. Again, these are my own terms, but they're a framework for thinking about the different responsibilities or different levels. Not everybody needs to be a data scientist, not even everybody needs to know how to run an aggression model, being able to make a bar chart and a line chart may be sufficient for a huge number of responsibilities in the everyday data or data literacy approach. And what that gets to is that there are terms that you may be familiar with. There's data literacy, which I just mentioned, and data fluency. There's also the terms quantitative or mathematical literacy, or maybe you're familiar with the term numeracy, which actually comes from a 1989 book by John Allen Paulos, which is called, "Innumeracy or Mathematical Illiteracy and its Consequences." But all of these refer basically to the ability to work with numbers, to work with data as a way of getting meaning and making decisions. So I'm going to be using these terms basically interchangeably. I know they're not exactly the same, but they have enough in common for us to use them productively. And with that out of the way, let me tell you what we're trying to accomplish in this course, or more precisely, what we're not trying to accomplish. This is what we are not doing. We're not trying to turn you into a data professional and, unless you're spending 100% of your time in data science, you don't need to be able to write your own code for neural networks, and you don't need to be able to find cube roots in your head. Besides, despite how often that kind of mental math shows up in the movies to indicate mathematical ability, it's really a separate, unrelated skill to thinking with data as any mathematician will tell you. So we're not going to do that. Instead, we're going to focus on the following things. We're going to see how you can learn when you can use data to answer questions. Data is helpful for an enormous number of things, but not everything. Second, we're going to talk a little bit about how to frame questions so that they can be answered with data. And we're going to present a small number of concepts and common methods that can help you find the answers to the questions that you want from the data that you have. And finally, how to interpret those analyses and apply those answers to the questions that you had in the first place. We're being very practical. It's a pragmatic exercise. Our goal is to use data to answer common questions without necessarily requiring a full on data science training. Again, data literacy and data fluency and the everyday elements of data for the vast majority of people are what we're focusing on, and trying to find a way to get direction and insight as quickly and as easily as possible.