Reproducibility

Open and Reproducible Work

It is good practice to conduct research in an open and reproducible manner, it improves trust in the research output and allows constructive feedback on the approaches taken.

Open Science

Open scientific research means being transparent about all steps taken in the analysis of data, from reading the raw data in, tidying it (renaming variables and deriving variables) through to running models, estimating parameters and making predictions.

Ideally work should also aim to adhere to the FAIR Principles and where software is developed, as is the case here where code for analysis is being produced, the FAIR4RS Principles too.

FAIR means research output is Findable, Accessible, Interoperable and Reusable. To help with this all work should on completion be added to the University of Sheffield Online Research Data Archive (ORDA).

Reproducibility

Reproducibility aids transparency and openness of research. By using scripted analysis and embedding code within literate documents we can streamline the work process and provide an accurate record of how we reached our conclusions from the data provided and the methods used. Further by version controlling the development of the scripts throughout their life-cycle we have a record of how the code has changed over time.

To aid with this the free open-source software R will be used to undertake data cleaning/preparation, statistical modelling and report writing. It integrates with the open-source scientific publishing system Quarto and as all scripts are based on text files they can be version controlled using Git. We will work collaboratively using the GitHub forge.

Because R packages evolve over time as bugs are fixed and new features introduced we will use renv to define a consistent set of packages that are used for analyses.

Getting setup with R, RStudio and Git/GitHub can be done following the excellent guide Let’s Git started | Happy Git and GitHub for the useR by Jenny Bryan.

References and Citations

Included in this repository is the file references.bib which is a BibTex formatted ASCII text file of citations related to this work. BibTex works well with Quarto and make including citations in work straight-forward (see documentation).