This the data and code repository for the multi-lab RRR of the Fischer et al 2003 to be published as:
Colling L. J., Szűcs, D., ... McShane, B.B. (2020) Registered Replication Report on Fischer, Castel, Dodd, and Pratt (2003). Advances in Methods and Practices in Psychological Science. doi: 10.1177/2515245920903079.
The raw data is located in the folder data. You can re-download the data from the lab's OSF pages by using the script DownloadAllData.r. Do to this, you will need to replace the string 'token_code' with your OSF token.
To ensure that the code in this repository compiles correctly there are two recommended options. Option 1 allows you to run all the code locally and it involves installing software on your computer. Option 2 allows the whole process to happen in the cloud.
Option 1 involves using Docker and the corresponding Docker container. To install Docker go to www.docker.com/products/docker-desktop. There is a pre-build container located on docker hub. To pull this container simply use the command docker pull lcolling/fischer-docker:canonical. This is the recommended method. Alternatively, if you wish to include additional packages, then you can build the container locally using the Dockerfile located in this repository. You can use the bash script builddocker.sh to do this.
Once you have pulled the docker image, you can run it using the command rundocker.sh. This will create a new RStudio instance accessible at http://localhost:8787. To login use the username rstudio and the password fischer-rrr. The R project file is located in /home/rstudio/manuscript/manuscript.Rproj.
As an alternative to using a local install of Docker, you can use Binder. Using Binder you can load up an instance of RStudio based on the lcolling/fischer-docker Docker image (including a few additions to make it work with Binder). To do this, simply click the launch badge below. This will spin up an instance containing the computational environment and the content of the this repository. Once the instance is running, the appropriate R Project (.Rproj) file will open and this file will be opened in the source pane. For best results, I recommend using Google Chrome.
All pre-processing and data analysis steps have already been performed so the manuscript can be built simply my loading the file manuscript_files/manuscript.Rmd and clicking the Knit button at the top of the script editing window. This will build the manuscript at manuscript.pdf.
If you wish to re-run the analysis then you will need to perform the following steps. For easy access files referred to in this guide, it is recommended that you load this file (README.md) into an editor window in RStudio. Files can then be accessed by clicking on the links below (cmd + click [mac] / shift + click [windows]).
Optional step: Download the data using DownloadAllData.r (the data is already contained in this repository so this step is optional).
- Run the pre-processing script at Pre_processing.Rmd.
- Run the data processing script at data_processing.Rmd.
- Run the data analysis script at analysis.Rmd
- This script will produce a number of files that will need to analysed using the online meta-analysis tool located at https://blakemcshane.shinyapps.io/mlmvmeta/. Instructions are provided by the analysis script.
- Run the script to produce the statistics at make_statistics.R
- Run the script to produce the figures at make_figures.R
- Compile the manuscript at manuscript_files/manuscript.Rmd
If you'd like to update the citation counts of the original paper in the manuscript then just run the command python get_citation_count.py > cites.txt at the terminal
The PsychToolBox Code for the main tasks are available in two repositories. The eye-tracker version is available here and the non eye-tracker version is available here
The main OSF page containing information about the pre-registration is available here
Information about the project and the participating labs is available here with more detail available here