Shiny Apps Team Successful at Poster Competition

UNIVERSITY PARK, Pa. — Dennis Pearl and Neil Hatfield are not ones to boast, but every summer and fall, BOAST is just what they do.

The BOAST (Book of Apps for Statistics Teaching) program was launched in 2017 by Pearl, a research professor of statistics, to provide undergraduate students the opportunity to develop coding skills while building creative applications that teach statistical concepts.

During the summer, students build their knowledge of an area of statistics they want to explore, create a concept, and develop their understanding and use of the popular R programming language used by statisticians and data analysts. The students’ concepts are then brought to life in Shiny Apps, a collection of web applications and modules that are built using R. Apps developed by students are reviewed and completed during the fall and deployed by the end of the semester.

Hatfield, an assistant research professor of statistics, said the experience is valuable for young students on multiple levels.

“Within the realm that we are working with our undergraduates, it's important for them for several different reasons,” he said. “First, to build an app to help someone else learn, you yourself must really gain a deep understanding of that concept; so one of the important things that comes out of this is that the students build more productive meanings for the ideas such as maximum likelihood estimation, sampling methods, or Bayes theorem, because they are having to think through, 'How do I support another person in understanding this?'”

Hatfield said that Shiny App development also gives the undergraduate students an opportunity to improve their R coding skills.

“Oftentimes, they don't necessarily get really strong feedback on how they're coding in a lot of our STAT classes,” he said. “They get that a little bit in STAT 184, which is our intro to R class; but beyond that, in their actual applied classes, they're sort of just left to code however they want.”

  Unique to the Shiny App development experience is the use of an R code style guide developed by Hatfield and Bob Carey, a programmer/analyst in the statistics department. Hatfield says following the style guide translates directly to real-world skills, as students going into industry will have to follow certain standards for their coding and documentation. To ensure students understand and meet those standards and build coding skills, Hatfield said he helps students refine, frame, troubleshoot, and debug their R code.

Another key component of Hatfield’s role is to encourage creative thinking.

“So, they're looking for, 'How do I create a game, how do I track participation across that game?'. I can help the students with that,” he said. “I also then kind of help with ideas for apps in terms of layout and even just the concepts that they're trying to tackle and share.”

For all the challenging work, Pearl, Hatfield, and six students get to “BOAST” a little extra this fall.

At the 2021 Undergraduate Research Poster Exhibition hosted by the Eberly College of Science, the BOAST Shiny Apps team — consisting of statistics undergrads Shravani Samala, Lydia Bednarczyk, Tina Tu, Jiayue He, Yudan Zhang, and Adam Poleski — was awarded second place in the mathematical sciences category.

Known at the competition as the R Shiny Research Group, the students presented their winning work from the summer, including apps exploring counting techniques, maximum likelihood estimation, probability applications, and chi square goodness-of-fit testing.

Of their success, Pearl said, "I'm really proud of our students who have taken on some particularly challenging topics to illustrate with web apps this year. This is a well-deserved honor that adds to the things they can boast about in job and graduate school applications."

Hatfield said the second-place award is significant.

“It's an acknowledgment of the hard work that they put in all summer,” he said. “Sometimes it is really challenging work because they are building apps from scratch. We have a couple of fantastic new apps, including the Counting Techniques app, where students got to design their own contextual situations in the digital space and sort of create this app around that.”

Speaking of the overall benefit of the BOAST experience, Pearl said, “The department strongly supports providing our undergraduate majors with research or other engagement opportunities like internships. The idea behind BOAST is to provide a paid summer program where students can improve their skills in R, have an experience in team-based reproducible programming, solidify their statistical content knowledge, and simultaneously generate resources for instructors across the entire undergraduate curriculum.  A key advantage is that we can readily supervise ten students in the program each summer rather than just one or two.  Also, providing funding to the students in the summer allows students with limited financial resources to take part — another important goal of the program.”

The Department of Statistics wishes to congratulate the BOAST program and Shiny Apps team on their success. More information about the undergraduate poster exhibition can be found on the Eberly College of Science website.

Instructors worldwide can also access the deployed Shiny App collection on the Eberly website, through the Shiny Apps pages.