 Hey folks, my name is Hannah Gunderman and I am a Research Data Management Consultant and Librarian with CMU Libraries, and also a very big video game fan. So in this first episode of Pixel DataScapes, I wanted to make a fun little video that talks about all of the data management lessons that we can learn through the video game Pokemon. So sit back, relax, and let's dive in. So first off, what is data management? I want you to think about all of the processes that you take to organize things in your day-to-day life, like your bookshelf, your spice rack, your kitchen cabinet, your record collection, so on and so forth. Now data management is basically applying those same concepts and goals that you have in those situations to your research data. So think about it, when you're organizing something like a bookshelf, you want to make sure that everything is tidy, you can find things quickly and retrieve them quickly, and you understand how things are organized. Same thing with data management in our research projects. Data management refers to all of the techniques we can use to make sure that our data are organized, easy to navigate, and understandable and reusable not only in the short term but in the long term. So in my job, I recognize that learning these concepts is not always the most fun. You know, we're not talking about all the fun data analysis that we can do in R and Python. We're not talking about the super cool data visualizations that we can do, but really data management is just as important. So when I started to think about, you know, what should this first episode look like? I thought, okay, I want to make data management fun, you know, something that's engaging to learn about, and in the game Pokemon, there's so many different things that really reflect on a very good data management practice. So why not combine the two? In this video, I'm going to describe three ways that Pokemon can teach us about data management. And these three things that I'm going to be talking about are code books, data documentation or documenting your data and metadata. Now these can be applied to any kind of data that you're working with, whether you are working with qualitative data, quantitative data, you're in the humanities, STEM, anything in between. So let's dive in. For those who don't know, Pokemon refers to a media franchise from Japan that has video games, card games, manga, television shows, and toys that revolve around a fictional world where humans become trainers for Pokemon, creatures who battle each other and gain new skills and power as they progress in their training. So let's dive in to our three examples of how we can learn data management from Pokemon. The first is code books. Code books are super useful for your research, especially if you are collecting and working with data and representing it in variables. It's always a good practice to name your variable something that is descriptive and makes sense based on the type of data represented by that variable, but it's also helpful to document what each of these variable names mean in a single place. In this place can be a code book. Now sometimes you might hear these referred to as data dictionaries. So as you can see here in this example, this is a code book that I have made for my own research, and you can see that I've listed the variable names associated with my data, what those variables represent, the measurement units, and allowable values, and a short description. This is going to help me remember in the long term all the important context of this data set, and will make it easier to share the data with others. Now a Pokedex is really just a code book for all of your Pokemon. As you can see here, a Pokedex is a really cool feature that is in every Pokemon game, that keeps track of every Pokemon that you've seen and or caught as you navigate this fictional world. Take for example this Pokedex entry for Sinistee, which is a ghost type Pokemon that looks like a really nice cup of tea. This Pokedex entry gives us a lot of helpful information about Sinistee, including their number, whether I as the player have caught them or not, their height, weight, and their description. Which reads, it absorbs the life force of those who drink it. It waits patiently, but opportunities are fleeting. It tastes so bad that it gets spat out immediately. Okay, that puts me right off my black tea that I'm drinking right now. Anyway, if you want to learn how to make a good code book, look to the Pokedex. It's a great way to keep all of the Pokemon and their important contextual information saved throughout your journey. And also a great way to keep track of your data. The second concept is writing documentation for your data. Now whenever we're working with data, we usually follow a step by step process of some kind, which might be something like collecting the data, tidying the data, looking for any data entry errors or missing values, running a statistical analysis, visualizing the results of this analysis, and then writing up these results in a paper. By the time that you get all the way to that last stage, you might not remember all the details of what you did to get to that point. And you know what, that is normal. We have so many things going on at the same time and we don't always remember why we chose to go with certain software over others or certain analysis techniques over others. But that's where documentation comes in. Writing down what you are doing as you work with data can benefit you not only now by keeping you organized, but also in the long term to help you remember what you did. So where does Pokemon come in here? Well, over the years as I've played the game, I've come to appreciate all the unique features that each Pokemon has that makes it different and special. I have a couple of favorite Pokemon that I think are absolutely beautiful. And when I describe them to people who aren't familiar with the game, I have to be really detailed and descriptive in order to convey what makes them so special to an audience that is unfamiliar. And one of my favorites is Moltres, a beautiful Pokemon that resembles the mythological bird known as a phoenix, rising from the flames. If I were to describe what this particular variation of Moltres looks like to a friend who doesn't play the game, I'd carefully list out all of their unique characteristics including the red and black stripes across their body, the red flames coming off of their wings and head, their blue eyes and determined expression and their large dark curved talons on each foot. This would likely give my friend a really good idea of what Moltres looks like without actually seeing the Pokemon in the game. Treat your data documentation process the same way. Assume that somebody one day unfamiliar with your research might want to reproduce what you did or just better understand the steps that you took to come up with your results. Be detailed and thorough just like you're describing the beautiful Moltres to a friend who's probably tired of me bringing up Pokemon in all of our everyday conversations. The final concept is using metadata to describe your data. Metadata is contextual information that can help us better understand a data set such as who created it, what was the time range of the data collection, what are the subjects or topics represented in the data, and what's the format of the data. Metadata is all around us such as the color of our eyes and the length of our hair. It's also everywhere in Pokemon and a great example are the different towns that your player navigates throughout the game. In each version of the game there are numerous towns which all have their own unique metadata such as the name of the town, the number of buildings, the types of businesses, and the natural features present such as trees, rivers, and oceans. For example, let's take a look at the town of Balanly in Pokemon Shield, the newest iteration of the game. This town has so many unique characteristics, in this case metadata, that describe it including all these unique glowing mushrooms in several different colors, the number of characters in the town, the number of houses, and the types of Pokemon present. These characteristics make each town different and unique from other towns in the game. And this descriptive metadata helps us capture those differences and easily make comparisons between different locations. Making sure you use metadata to describe your data is one of the most impactful things you can do in a research project to make sure that your results are understandable to you. And no matter what field you're researching in, there's a great metadata standard out there for you. We've got some great folks on staff here at CMU Libraries who can help you find a metadata standard that works best for your research. And the person that I would probably recommend first is Angelina Spotz, who is our metadata specialist, and I've left their contact information below in the description box. I hope this video has been entertaining and helpful for your research process here at CMU. All of us at CMU Libraries are here for you every step of the way when you're working with data and communicating with data. If there was a concept that I talked about today that you'd like to explore more, you can feel free to send us an email at dataatcmu.libansers.com and we'd be happy to start the conversation. So thank you for watching the first episode of Pixel Data Scapes and we'll see you next time. Stay safe, everybody.