19 Conclusion: Where to next?

Abstract This chapter contains bigger-picture advice about how to apply the capabilities and knowledge one has developed with respect to data science education, with an emphasis on the importance of continuously connecting one’s data science-related work to the language, problems, and types of datasets encountered in education; collaborating and building trust with others;, and—–on occasion—–taking strategic breaks by stepping away from the keyboard to help solve problems.

To start closing our journey together, let’s recap what we’ve learned so far. When we started writing this book, we set out to create a learning experience that had recognizable education examples as its foundation. We used these examples to explore the role of data scientists in education. Building on that context, we introduced basic R tools. In the analysis walkthroughs, we learned to apply data science techniques to datasets and scenarios we’ve seen in our education jobs. Our goal was to help you learn data science using datasets, functions, language, and analytic approaches that you’ll keep using in your own education job. Finally, we discussed using these technical skills to positively influence how your education organization uses data.

After all that, we hope that you feel more prepared to take on the data-related questions and problems that matter to your students. We want you to feel excited about choosing your next steps. But there’s one more thing we’d like to share about data science in education: how do we keep our efforts going in the long-term.

As exciting as the promise of data science can be, over time there are inevitable bumps in the road that all educators encounter. As you master doing data science with education datasets, your learning challenges will evolve beyond the problems of coding syntax and into the realm of larger questions about education systems. For example, you may take an interest in new coding or statistics techniques but struggle to find the right application in your education job. Or you may find yourself working alone on a data project when you need a collaborator as a thought partner. Or maybe you’re wondering how a project you’re working on aligns with your values as an educator.

When these questions come up, it can be inspiring and reinvigorating to reflect on the bigger picture. Try thinking a little less about what you learn and a little more about how you learn, why you learn, and the ways your work can positively affect students and staff in your education system. Think of this as a strategic move to bring back the excitement and hope you feel when you solve problems that truly make the lives of your students better. In this final chapter, we’ll discuss ways to think about your journey that ground you in your service to people learning in your education system. Let’s kick off the next stage of your learning!

19.1 Learn in the context of education

The mental models and technical skills we learn are separate from the places and people we practice them with. If you intentionally learn new skills in the context of your daily work, you’ll naturally gravitate towards the skills that are right for the problems that you are trying to solve. It makes sense for an auto mechanic to learn the science of diagnostics while working on automobiles. It makes sense for a surgeon to learn about anatomy while mastering surgical methods. It makes sense for a farmer to learn about the ecosystem while growing crops. In all these examples, the data scientist concerns themselves with mastering skills in service of solving meaningful problems for people.

We’ve already started learning the basics of data science in education by introducing tools like R. Mastering R will help you analyze data at scale. It will also help you share your work while your organization’s data practices grow. But as these skills develop and become muscle memory, your time and attention can focus more and more on what this data tells us about the students and staff that actually generated the numbers to begin with. This is why we believe so strongly in teaching data science using language, problems, and datasets commonly found in education. It helps you make the cognitive jump from simply working on datasets to working on datasets in service to students.

We hope that you’re excited to take what you’ve learned and share it with others. Remember that what you share is not just a way to analyze data, but a way to contribute to someone’s school experience.

And speaking of sharing, let’s talk about one of the most powerful creative tools we have available to us: each other.

19.2 Learn to collaborate with others

In the last section, we talked about data in the service of people. But working with data is as much about working with people as it is about working for people. To understand why collaboration in data science is so important, we first have to see data analysis as a fundamentally creative process. We don’t mean this in the same sense that art is creative. First, data science is creative because practitioners create a process that extracts meaning from data. Then second, they create output like writing, visualizations, and conversations that convey this meeting to an audience.

In most creative endeavors, collaboration is the magical ingredient that evolves an individual idea into something truly unique and responsive to the needs of an audience. Daniel Kahneman and Amos Tversky brought the world new knowledge about cognitive biases (see Kahneman -Kahneman (2011)). Ben Cohen and Jerry Greenfield collaborated to create Ben and Jerry’s Ice Cream. Data journalism like the kind you find at FiveThirtyEight and The Economist uses collaborative work to produce many visualizations and analyses that consistently deliver high-quality information.

When we started writing this book together, we knew early on that the best product would come from a truly collaborative experience. When you work with others to write words, write code, and think analytically, you learn practices that create inspiration and excitement about the ideas you want to bring to life. Here are some lessons that we learned in our journey:

  • Building trust with your collaborators will lead to productive experimentation with new ideas. Build trust slowly with your collaborators by actively listening to feedback and taking risks with new ideas from your teammates.
  • Adopting an experimental mindset will maximize the opportunities to fail fast and find the solutions that are right for the problem. If you and your collaborators brainstorm an idea that feels promising but ambiguous, try it out and evaluate the results. Do this together often.
  • Asking for feedback and giving feedback when asked will lead to a refined end product. Ask your collaborators regularly for their reaction to a visualization you’ve made or a report you’ve written. Does their feedback provide evidence that you’ve made what you intended to?
  • Starting a draft of a project then releasing it to a collaborator is an exercise that builds trust. During our editing process, Josh reminded us of his collaborative spirit by saying, “I no longer consider anything I’ve written in this book mine. Change anything you want!”
  • Collaboration is contagious. There’s a really easy way to make a work environment more collaborative: approach someone and invite them to collaborate. You might be surprised at how being an active collaborator inspires similar behavior from others.

19.3 Learn every time you code

Whether you’re working on a solo project or on a collaborative project, you’ll often find that completing it requires learning something new. Learning requires you to get comfortable with not knowing how to do something because it liberates you from the pressure to know all the time. And when we don’t feel pressure to be perfect, our minds are free to enjoy the challenges of learning.

When some of us first started learning to program in R, the number of things we didn’t know was glaring. We’d never typed a line of R code in our lives. The thought of writing code that worked, much less using it to do data analysis, felt like a distant goal. After lots of practice and patience, we find ourselves writing code and doing data analysis every day in our education jobs. Yet we still have a long list of things we want to learn so we can push for new ways to understand the lives of our students. Our vocabulary of R syntax has grown, but we still regularly enjoy the experience of learning a new function, discovering a new package, or learning about an analytic workflow. The difference between our early learning experiences with R and more recent experiences is that we’ve embraced learning as a necessary part of enjoying this craft.

It’s daunting to begin learning a new function, concept, or statistical technique. We know it will require sustained discomfort, trial and error, and some frustration before we experience the sweet thrill of a well-executed code chunk. Have you ever noticed how hard it is to get started? But simply beginning the learning process is like strengthening a muscle through exercise: it’s really difficult at first but with repetition, patience, rest, and kindness towards yourself, it gets easier.

So we encourage you to just start. Set a small goal to open up that book about machine learning you’ve been avoiding and get through that first paragraph. Or fire up RStudio and copy and paste that first code chunk from the GitHub repository you’ve been trying to understand. Or run the first example from the documentation of the package you’ve been trying to learn. Trust us—you get used to it and, eventually, you’ll start to enjoy all that comes with learning.

But even the most motivated of us can’t sustain high effort and challenges indefinitely. When you’ve hit a wall trying to learn a new concept or you’ve tried to fix your code one too many times, it may be time to take a strategic break.

19.4 Learn to take strategic breaks to help solve problems

Taking breaks is one of the most strategic moves you can make when you’re trying to break through to the other side of a programming challenge. And it’s not just true of R programming. Ryan talks about one of the first times he saw taking breaks in action:

I grew up in the 1980s and 1990s when completing Super Mario Brothers was a monk-like endeavor. I spent hours helping Mario navigate across green-colored plumbing that unexplainably stuck out from the ground. I practiced combinations of jumps and runs to avoid Koopas to eventually reach the end of a two-dimensional level.

Looking back, there were moments when playing this game reached levels of frustration comparable to pressing an impossibly small Lego piece into a complex Lego structure that fell apart in my hands. When these frustrating Mario moments came, there was only one strategy: attempting the challenge over and over again until the inevitable throwing of the controller.

After throwing the controller, I’d engage in some other activity. I’d take a nap, go outside, hang out with my sister, or watch TV. Strangely, when I returned to the game console to play again, I’d often progress through the same frustrating challenge on the first or second try. Some moments I succeeded so quickly it was hard to believe the level was ever a challenge to begin with.

I learned a valuable lesson from indulging in this exercise repeatedly for many years: during moments of frustration, the mind and body need a break. Taking breaks is a functional activity. It gives you time to synthesize all the learning that happened during repeated attempts to solve a problem. During the breaks, the mind and body work to replenish energy to take on problems again. Taking strategic breaks is not failure, it’s the mark of a professional who understands the most efficient way to get to viable solutions.

So when you need it, take a break and live to code another day. We need as many data scientists working in different corners of education as possible, and we don’t want to lose you to burn-out!

19.5 Learn more meaningfully by knowing your why

Our last bit of advice for staying connected to the bigger picture: make time to regularly reflect on why you’re doing this work to begin with. Doing data science in education without reflecting on the cause you’re trying to positively affect can make you feel a little like you’re just spinning your wheels. Fortunately, working in education has a common built-in purpose: to provide the highest quality learning experience for students with the resources you have available.

To be clear, getting familiar with the “why” of your work is an ongoing process. As we learn and grow, our motivations evolve. But still, better to have this evolution be an intentional and self-aware journey. Here are some reflection questions that are useful to prompt the kinds of thinking that activate your “why”:

  • Think of what your own learning experience was like. Were there things you wish were different? Were there positive experiences you hope more people will have?
  • Is there a teacher or school leader that you’ve worked with or that positively influenced your life as a student? What was it about them that you’d like to see more of in schools?
  • Is there an education-related topic, like diversity, special education, or curriculum and instruction, that you have a natural passion for?

Most people don’t have immediate answers to questions like these. It’s good enough to just ask them regularly and reflect on the thoughts and feelings that come up. Finding your “why” and being able to talk about it is more like an exploration and less like a singular “aha” moment. But we believe asking these questions is a way to intentionally design meaning into your data science work. When there is a clear purpose and personal connection to why you use data science in education, you’ll be reminded that your work affects the lives of learners, sometimes in profoundly meaningful ways.

That brings our time together to a close, but only for a short while. We hope to see you in the “data science in education” community, offering value, learning from others, bonding over challenges, and inspiring each other to the best we can for our students.