Sharing and organizing research products as R packages

Area of Science:

• Psychological Science
• Computational Science
• Data Science

Background:

• A growing consensus emphasizes the importance of open data and reproducible code in research.
• Sharing research assets effectively requires balancing reproducibility, permanence, and accessibility.
• Current methods for sharing research data and code may not be easily adopted by busy researchers.

Purpose of the Study:

• To propose a practical and effective solution for sharing research assets that maximizes reproducibility, permanence, and accessibility.
• To advocate for the adoption of the R package standard as a robust method for curating and communicating research data and code.
• To introduce a suite of R resources and a specific package ('vertical') designed to streamline the creation of reproducible research workflows.

Main Methods:

• The study proposes the R package standard as a format for organizing research assets and associated metadata.
• It discusses extensions to the R package standard to enhance its utility for reproducibility and open access.
• The authors introduce and detail the functionality of the 'vertical' R package, integrated within the RStudio environment, to facilitate reproducible workflows.

Main Results:

• The R package standard, with proposed extensions, provides a software-agnostic format for research assets and metadata.
• This standard facilitates the creation of persistent, accessible, and reproducible research materials.
• The 'vertical' R package and associated R resources offer a practical stack for researchers to manage reproducible workflows from experiment to manuscript.

Conclusions:

• The R package standard offers a viable solution for enhancing the sharing of research assets, promoting reproducibility and open science.
• The proposed tools and workflow integrations can significantly increase the openness and reproducibility of scientific research, particularly in psychological science.
• Adoption of these methods can lead to more robust, verifiable, and extensible scientific findings.