The TIER Documentation Protocol provides instructions for assembling a set of electronic files that document all the steps of data processing and analysis you conduct for an empirical research paper.
The documentation specified by the Protocol contains all the data, computer programs, and explanatory information an independent researcher would need to be able to replicate the data processing and analysis you conducted for the project and to reproduce exactly all the results reported in your paper.
The instructions presented in this repository are written for users of R. In a few places, they use R-specific terminology. For example, we refer to command files as scripts, and their names are followed by the .R extension. But the R-specific terminology that appears in these instructions can be easily translated to any of the major statistical packages (such as SPSS, SAS, Stata or Matlab) or other programming languages.
To get started you can fork and then clone this repository which will create a copy of the folder structure recommended in the Project TIER protocol.
Below we describe how to organize your analysis according to the Project TIER protocol, i.e. which components of your analysis should go into which folder.
Your repository should have the following hierarchy of files and folders:
- An electronic copy of your complete final paper
- The
README.md
file for your repository - Original Data and Metadata -
original-data-and-metadata
- Original Data -
original-data
- Metadata -
metadata
- Metadata Guide -
metadata_guide.md
- Supplements -
supplements
- Metadata Guide -
- Original Data -
- Processing and Analysis -
processing-and-analysis
- Importable Data -
importable-data
- Command Files -
command-files
- Analysis Data -
analysis-data
- Importable Data -
Contents of these files and folders are described in the README
files
within these folders.
The README.md
file in the top hierarchy of your repository (this
file) gives information about all the other files included in the
documentation for your paper. In particular, the README
file should:
- state what statistical software or other computer programs are needed to run the command files.
- explain the structure of the hierarchy of folders in which the documentation is stored, and briefly describe each of the files included in the documentation.
- describe precisely any changes you made to your original data files
to create the corresponding versions saved in your
importable-data
folder. - give explicit, step-by-step instructions for using your documentation to replicate the statistical results reported in your paper.
The README should be a Markdown document so that it can be
rendered properly on GitHub, and any changes can be tracked. It should
be named README.md
. This file should be stored in the top level of
your repository.