Commit Plan
Apr-07
: In this lesson, we will finalize a reproduction study analysis plan.
Goals and Expectations
The overarching goal is to finalize the analysis plan.
A practical checklist:
- Create repository and share ownership with partner
- Fill out top-level readme.md
- Update LICENSE
- Complete one metadata readme file for each data source
- Save available data into the repository
- Update the data_index.csv
- Choose R or Python analysis plan template and start fillling out
- Copy (or import with code) content from top-level readme and metadata readme’s into the analysis plan
- Review of the paper’s context, significance, and methods
- Add and justify any planned deviations
- Discussion of implications: what if the reproduction succeeds at generating consistent results? What if it does not?
- Identification of required computational environment / software packages
- Optional: Outline or even draft code blocks
- Knit the analysis plan to
html
and save in your docs
folder with suffix _plan.html
- Update the
procedure_index.csv
- Create a version
release
of your compendium on GitHub prior to analyzing any data and commit a version of the repository prior to analyzing data. Version releases shown starting at about 10:00 in this video
- Enable GitHub pages on the repository
- Draft a very short blog post linking to your proposed study release and highlighting your interest in the study.
As part of your preparations, you may want to identify and run some tutorials or vignettes that illustrate the methods you will need for the study.
Spin this up in separate Learn-
repositories as we have done with these:
Reading
- B.A. Nosek, C.R. Ebersole, A.C. DeHaven, & D.T. Mellor, The preregistration revolution, Proc. Natl. Acad. Sci. U.S.A. 115 (11) 2600-2606, https://doi.org/10.1073/pnas.1708274114 (2018).
Main Page