How do we help students carry out data analysis workflows that are comprehensible and reproducible? The 2018 NASEM “Data Science for Undergraduates” report enunciated the importance of workflow and reproducibility as a key component of data acumen. “Documenting and sharing workflows enable others to understand how data have been used and refined and what steps were taken in an analysis process. This can increase the confidence in results and improve trust in the process as well as enable reuse of analyses or results in a meaningful way” (NASEM 2019, page 2-12).
Leaflet for mapping Maps are an important way of displaying data. The leaflet package in R provides access to the Leaflet Javascript libraries (http://leafletjs.com), an open-source mechanism to create interactive maps. The leaflet package (https://rstudio.github.io/leaflet/) provides an interface within R that allows for composing maps using map tiles (e.g., from OpenStreetMap, https://www.openstreetmap.org/#map=5/38.007/-95.844) that can be annotated with markers, lines, popups. Here’s a simple example where data from higher education institutions from within the Five College Consortium in Western Massachusetts is mapped.
The National Academies of Science, Engineering, and Medicine Roundtable on Data Science Postsecondary Education was convened in 2016 to work to develop a coherent vision for the emerging field of data science. The Roundtable, which consists of representatives from industry, government, and academia meets quarterly for a day to discuss best practices, ways to support the growing community, and approaches to help advance data science education. Full information about the roundtable including narrative summaries of past meetings can be found at https://nas.
Calls for Diversity Data science is made up of not only sets of tools, methods, and problems to solve, but also actual people who make up the statistics & data science community. The National Academies Report on Data Science for Undergraduates (see previous blog post at: https://teachdatascience.com/nasem) includes a section on “Ensuring Broad Participation” which reiterates the importance of creating an inclusive community where all views are heard and supported.
The American Statistical Association has placed a priority on how best to teach statistics and data science. The Guidelines for Assessment and Instruction in Statistics Education (GAISE) reports have served a key role in guiding instructors and institutions in their pedagogical choices. Two GAISE reports have been written: one focused on statistics at the PreK-12 level and another, revised in 2016, focused on college level courses. In this GAISE blog entry we focus on the college report.
As an instructor teaching data, it is often difficult to explain the world the students will be joining (industry) given the experiences of the instructor (academia). One way to bridge the two worlds is to peek into the world of data science outside of academia and then tell your students about it. Hilary Parker and Roger Peng’s podcast, Not So Standard Deviations provides glimpses into data science challenges, obstacles, opportunities, and solutions in the real world.
Watching an expert work through a data analysis using the tidyverse Teaching Data Science is challenging since it involves teaching the entire data science analysis cycle. While it’s helpful for students to experience this process, they can often feel at sea in terms of the decisions they need to make and the iterative process of exploration, modeling, summarization. We’ve been using the data science cycle promulgated by Hadley Wickham and Garrett Grolemund (both from RStudio) that was published in their excellent book: R for Data Science, https://r4ds.
What is the Tidyverse? The tidyverse is a coherent system of R packages for data wrangling, exploration and visualization that share a common design philosophy. These packages are intended to make statisticians and data scientists more productive by guiding them through workflows that facilitate communication, and result in reproducible work products. Unpacking the tidyverse, all that it means and contains, could easily take a dedicated book or blog in itself.
What are Projects? RStudio Projects are a mechanism for keeping all the files associated with a project together in one place – data, R scripts, results, figures, reports, etc. Projects are built in to the RStudio IDE, and for good reproducible workflow, all projects should start by creating a Project. Why RStudio? It goes almost without saying that as a group we have moved completely to the RStudio interface to R.
What is R Markdown? Straight from RStudio’s wonderful tutorial, R Markdown is an authoring framework for data science. An R Markdown file is a plain text file with three types of content: code chunks to run, text to display, and metadata to help govern the R Markdown build process. Put simply, R Markdown is an exciting new reporting medium that seamlessly integrates executable code and expository text. By including data work, code, and analysis narrative into a single document, R Markdown provides a fully reproducible vehicle for data science projects!