This repository contains the code and data for replicating the results of a messaging experiment conducted by the Animal Welfare Action Lab (AWAL) with the support of the Reducetarian Foundation.
The working paper, pre-registration, and surveys for this study are available at: https://osf.io/44z7j/.
We welcome contributions to the existing set of analyses (see TODO.md
for a list of analyses that have not been completed yet). If you wish to contribute, follow the instructions here to fork the repository, make changes, and submit a pull request.
Citation:
Macdonald, BNJ, KD Caldwell, and GD Boese. 2016. The effects of "reduce" and "eliminate" appeals on individual meat consumption. Unpublished manuscript. Available at: https://osf.io/nxrx3/.
- Download the entire repository.
- To replicate the data cleaning (i.e. converting raw data files in
raw-data-shareable
into clean data files incleaned-data-shreable
), open a command line session and run:$ bash clean.sh
- To replicate the data analysis, open a new command line session and run:
$ bash analysis.sh
This will run the R scripts in order and reproduce the main results contained in the results
, tables
, and figures
directories.
01-wave1-cleaning.R
cleans the wave 1 Qualtrics data and saves it as a csv file.02-blocking.R
was used for block randomized assignment to treatment.03-wave2-cleaning.R
cleans the wave 2 Qualtrics data and saves it to a csv file.04-wave3-cleaning.R
cleans the wave 3 Qualtrics data and saves it to a csv file.05-merging-all-waves.R
merges the data from waves 1, 2, and 3 together and does some additional cleaning.06-main-analysis.R
contains the analyses for producing the main treatment effects.07-subgroup-analyses.R
contains sub-group analyses examining how the treatment effects vary across groups in the study sample.08-alternative-analyses.R
contains analyses for examining treatment effects on other outcomes of interest.
cleaned-data-shareable/all_waves_cleaned.csv
contains the cleaned data output by05-merging-all-waves.R
that is used to produce the main results.raw-data-shareable/
contains the raw unprocessed data.- see
other-scripts/variables.r
for a description of variables in the cleaned data files. Seevar-dicts/
for the interpretation of variable names in the raw data files. - all MTurk IDs have been replaced with hashed IDs to ensure anonymity.
var-dicts/
contains variable dictionaries for the raw data files.functions/
contains functions that are called in the main R scripts.other-scripts/
contains additional R scripts that are called in the main R scripts.