worcshop-book

Writing A Reproducible Manuscript in R with WORCS

This repository contains materials for the single-day workshop 'Writing a Reproducible Paper in R with WORCS' at Utrecht University.

Workshop Description

Open Science is becoming increasingly popular and relevant, and a world of opportunity is opening up to make your work fully reproducible. This is not without its challenges: best practices for reproducible science include a number of tools that you may never have used or even heard of before - Are you using version control? How are you managing your dependencies? Are you writing your manuscript as an executable document?

For those who would like to get started with an open and reproducible workflow, without dealing with a mountain of new tools and platforms, we introduce WORCS, a Workflow for Open Reproducible Code in Science. The workflow is written for R, but you do not need to have prior programming experience to join this workshop. Having the motivation to step out of your comfort zone — and into a new one — is the most important prerequisite.

WORCS is an R package that takes you from data to published paper in a single streamlined workflow, making the entire process of your analysis, up to the submission of your manuscript, reproducible. The WORCS workflow optionally facilitates pre-registration, sharing your code and your data (safely!), and the submission of a pre-print.

This workshop will be taught by Research Data Management Support in collaboration with Caspar van Lissa, author of the worcs package and the WORCS workflow. We will give you an overview of the workflow and introduce you to its use. You will create your first reproducible project by going through all steps of the workflow yourself.

Upcoming Workshops

This is a pilot workshop. Upcoming workshops will be featured on RDM Support's trainings & workshops page and the university's Events page, in addition to the Intranet and various (internal) newsletters.

Setup & Installation Requirements

Please work through the Setup & Test instructions prior to the workshop. If you run into issues during setup and installation, you can contact your instructors to help resolve them.

Schedule

This workshop will be given online. Participants will therefore do most exercises independently, using the step-by-step guides provided. Instructors will be available for 1:1 support during these times, and reconvene in the main channel at stated times.

Time	Activity
9:00	Walk-in, Tech Support
10:00	Introductions
10:15	Introduction to WORCS by Caspar van Lissa
10:45	Explanation of workshop program and setup
11:00	Phase 1: Study Design
11:45	Recap & Questions
12:00	Lunch Break
12:30	Phase 2: Writing & Analysis
13:15	Recap & Questions
13:30	Phase 3: Submission & Publication
14:15	Recap & Questions
14:30	General Discussion
15:00	End

Data

You are encouraged to use your own data, code, and manuscript text during the workshop.

If you do not have your own data, you can use Allison Horst's Penguin dataset.
You can use this data by entering the following code in the prepare_data.R script.

In case you need some code to run with the Penguin data, you can use some of the example code on the palmerpenguins R package website.

install.packages("remotes")
remotes::install_github("allisonhorst/palmerpenguins")
data <- palmerpenguins::penguins

Alternatively, you can download one of these versions of the data, and load them yourself:

• An SPSS file (penguins.sav)
• An Excel file (penguins.xlsx)
• A CSV file (penguins.csv)

Workshop Program

Links to a step-by-step guide for each part of the workshop are provided below, taken from the WORCS Workflow Vignette.

Subject	Link	Additional information
Phase 1: Study Materials	step-by-step guide	-
Phase 2: Writing & Analysis	step-by-step guide	citing references in WORCS
Phase 3: Submission & Publication	step-by-step guide	-

If you're new to R Markdown, here's a handy cheatsheet!

Notes for cautious researchers

Some researchers might want to share their work only once the paper is accepted for publication. In this case, we recommend creating a "Private" GitHub repository in Step 1, and completing Steps 13-17 upon acceptance by the journal.

Additional Resources

• This paper in Data Science introduces the WORCS workflow, explains the underlying tools, and illustrates how the worcs package can be used to create a new project that follows the workflow. You can cite WORCS, both the package and paper, using the citation below.

Van Lissa, C. J., Brandmaier, A. M., Brinkman, L., Lamprecht, A., Peikert, A., , Struiksma, M. E., & Vreede, B. (in press). WORCS: A Workflow for Open Reproducible Code in Science. Data Science, 2021. Data Science, vol. 4, no. 1, pp. 29-49. DOI: 10.3233/DS-210031.

• This presentation from an earlier workshop introduces WORCS and the concepts + tools its based on.

• Caspar van Lissa has recorded the following videos introducing WORCS and showing WORCS in action.

• The worcs GitHub repo & package documentation

• R Markdown cheatsheet

Sample WORCS projects
For a list of sample worcs projects created by the worcs authors and other users, see the README.md file on the WORCS GitHub page. This list is regularly updated.

Acknowledgements

This workshop was developed by Research Data Management Support at Utrecht University, in collaboration with Caspar van Lissa. The workshop repository and documentation was set up by Neha Moopen and Bianca Kramer, based on an earlier workshop by Barbara Vreede.

Image attribution
The Git Logo by Jason Long is licensed under the Creative Commons Attribution 3.0 Unported License. The OSF logo is licensed under CC0 1.0 Universal. Icons in the workflow graph are obtained from Flaticon; see detailed attribution.