This is the repository for the following article: Vogel, D., & Homberg, F. (2020). P-Hacking P-Curves and the PSM–Performance Relationship: Is there evidential value? Public Administration Review. Advance online publication. https://doi.org/10.1111/puar.13273
A **long-time archived** version of the data and code is available at https://zenodo.org/record/3951892
There is a **Code Ocean Capsule** for long-term computational reproducibility: https://doi.org/10.24433/CO.5587415.v1. See below for more information.
``Docker-Capsule.zip`` contains a **Docker image** that can be run locally and ensures long-term computational reproducibility. See below for more information.
### Reproducibility
<strong> Using Code Ocean </strong>
You can easily reproduce our results using the free service of [Code Ocean][1]:
1. Open the paper's Code Ocean Capsule: https://doi.org/10.24433/CO.5587415.v1
2. Log-in to Code Ocean
3. Click *Re-Run*
4. Code Ocean runs a virtual environment and reproduces the results.
You can inspect the *R* code by clicking on the `code/Analysis_CodeOcean.R` file in the left panel (see below for an explanation of the file structure).
You can also run an interactive version using Code Ocean.
<strong> On your own machine (using your own *R* installation) </strong>
If you want to reproduce our results on your own computer you need *R* (we used version 4.0.2) and RStudio.
1. Download the full OSF Storage of the Project (including `.here` and the folders `code`, `data`, `results`
2. Make sure that you installed all required packages (see below for a list of required packages)
3. Open `code/Analysis.R` in RStudio.
4. Run the code
<strong> On your own machine (using Docker) </strong>
1. Install Docker Community Edition
2. Download the `Docker-Capsule.zip` file
3. Open a Terminal and navigate to the folder where the file was saved
4. Execute the following command:
```shell
docker load --input Docker-Capsule.zip
docker run --rm \
--workdir /code \
--volume "$PWD/data":/data \
--volume "$PWD/code":/code \
--volume "$PWD/results":/results \
p-curve bash run
```
<strong> File structure </strong>
The repository consists of some files in the root folder and three folders.
* `.here` is an empty file that helps the `here` package to set the right working directory in *R*.
* `/code` contains the analysis script:
* `Analysis.R`: main file containing the analyses.
* `/data` contains the data of the paper in a CSV file:
* `Meta-Analysis.ods`: Raw data for the meta-analysis
* `p-curve.ods`: Raw data for the p-curve analysis
* `/results` empty folder that stores the result of the analyses.
<strong> Required R packages </strong>
- stringr
- poibin
- metafor
- readODS
- here
- tidyverse
It is advised to use `R` version 4.0.2 with the `checkpoint` package and set the date to 2020-07-17 to ensure that the package versions are comparable:
```R
library(checkpoint)
checkpoint("2020-07-17")
```
[1]: https://codeocean.com
