Contributing to Scientific Reproducibility: Using R Markdown and Knitr to Create Reproducible Research Documents

Reproducing findings is essential to maintaining the legitimacy of science. However, recent research has shown that results of many research studies are difficult or impossible to reproduce (Yaffe, 2019). This replication crisis undermines scientific credibility and affects social scientists and the community that relies on these results. Tools to increase transparency and guarantee a reproducible workflow in research are now available. This workshop aims to teach attendees how to make their research more reproducible and shareable using R Markdown.

The R Markdown and Knitr packages in R Studio (Allaire et al., 2021; Xie, 2014, 2015, 2021; Xie et al., 2018, 2020) provide an environment that enables a workflow allowing code, output, and regular text to coexist in only one reproducible document. Therefore, researchers can share their work and provide necessary details about the study procedure and data. Learning how to build research documents in R Markdown will contribute to the transparency of research and assure the methods and results can be reproduced by others. The workshop will use a combination of lecture and practice to produce research documents that include text, figures, and tables. Attendees will know how to build the sections of the research paper, include figures, create descriptive statistics tables, and include in-line citation and full references lists. Additionally, an introduction to the rUM package (Balise & Odom, 2021) will also be covered, given it shortens the time it takes to build academic scholarly papers.

This workshop will begin with a lecture format to raise awareness about the reproducibility crisis and the factors that contribute to it. Then attendees will access RStudio.Cloud to recognize and manipulate all the features included in a R Markdown document. This will be followed by familiarizing with the introduction of the packages and utilities to produce the document (rUM), wrangle the data (tidyverse), create the tables (table1), generate the figures (ggplot2), and reference (.bib). Finally, attendees will follow the workflow of the construction of the reproducible document on RStudio.Cloud and work on examples and exercises.