What is R Shiny? shiny is a powerful and flexible R package that makes it easy to build interactive web applications and dynamic dashboards straight from R. These apps can be hosted on a standalone webpage or embedded in R Markdown documents. Not only does shiny allow you to build these web apps from R, but it enables their construction using only R code. Knowledge of HTML and web development is not required at all, though it can be used to enhance your apps in numerous ways.
As educators, it is exciting that our course enrollments are up and students are excited about data science topics, models, software, and careers. It is also validating that when they graduate, our students are able to support themselves doing interesting and engaging work. However, it can be sometimes disheartening to realize how many of our data science students use their skills to maximize the number of times viewers click on ads.
As statistics educators, it is often easier to focus our teaching on methods instead of communication. And while many of us understand the value of good communication, actually teaching it is difficult and outside of our comfort zone. There has been quite a bit of work done on the science of visualization (e.g., the Grammar of Graphics by Wilkinson). There is general consensus that teaching students to communicate using visualizations is of paramount importance (see recent blog entries: National Academies Report on Data Science and GAISE).
When people ask about how to get their students engaged with R in their introductory statistics and data science courses we offer three pieces of advice: keep it simple (discussed in the “Less Volume, More Creativity” blog entry) engage students to provide peer-tutoring and drop-in office hours to assist with questions and coding to complement class and office hours (at Amherst College this is coordinated by the Statistics and Data Science Fellows) have students use a dedicated server to access R Slide credit: Mine Çetinkaya-Rundel
Pair programming is a technique from software development where two programmers work in tandem to code. One is designated the driver, responsible for typing, while the other, often called the navigator or observer reviews the code and provides a high-level overview of the task. Photo credit: Esti Alvarez Pair programming has been thought to lead to better code, more enjoyable coding, and higher productivity, with some research findings supporting those conclusions (see some of the references at the end of this entry).
In 2017, Jenny Bryan and Hadley Wickham published the “Practical Data Science for Stats” PeerJ collection. (The papers were also published in a special issue of The American Statistician.) The “Practical Data Science for Stats” Collection contains a series of short papers focused on the practical side of data science workflows and statistical analysis. There are many aspects of day-to-day data analytical work that are almost absent from the conventional statistics literature and curriculum.
In 2016, GAISE enunciated the importance of multivariate thinking and technology when teaching introductory statistics and data science courses. A big challenge is how to do this using R and RStudio without running into cognitive overload with our students. The mosaic package was created by Randall Pruim, Danny Kaplan, and Nicholas Horton with the goal of introducing a Less Volume, More Creativity approach to introductory statistics that could simplify the use of technology.
Although an agreed upon definition of data science is hard to come by, there is clear consensus that statistics plays a key role in the foundational knowledge of anyone working with data. One important aspect of statistics is understanding of the inferential process that allows claims to be made about a population from a dataset. Most Introductory Statistics courses and textbooks spend substantial time presenting statistical inference as a way to generate p-values and make claims (or not) about a research hypothesis.
GitHub Classroom If you have been reading along in the blog, you’ve noticed the last two entries describing GitHub and GitHub in R. And certainly, we continue to advocate teaching students to use GitHub as an integral part of their data science workflow. And GitHub may be the perfect place to store student projects either as public or private repositories. But using GitHub to navigate a dozen homework assignments with 50 students can become logistically difficult.
Once you get the hang of using Projects in RStudio, you may be inclined to collaborate with others on the same project. If so, you will want to set up a Project that links directly to GitHub. By having your project on GitHub (and regularly saving it / updating it on GitHub), your collaborators will always have access to the most up to date analysis information. Previous posts have described working with R Projects and working with GitHub.