Main content

Home

Menu

Loading wiki pages...

View
Wiki Version:
This is the OSF project page for the paper titled "In COVID-19 public health messaging, loss framing increases anxiety without concomitant benefits". This study (henceforth referred to as PSACR-001) is one of three studies in a COVID-19 study package that was coordinated in collaboration with members of the Psychological Science Accelerator (https://psysciacc.org/). More information about the study package can be found here: https://osf.io/3g7vq/. Code ---- **psacr001_datacleaning.Rmd** psacr001_datacleaning.Rmd merges and cleans data from three .csv files, which the user should store in a local folder titled "Data". Original versions of these files are included in the Data folder on this OSF page, but updated version of the files can be downloaded using the links below: - study1_data.csv: contains responses to the study 1 survey (download here: https://osf.io/ctsrk/) - study1_country.csv: contains participant-level demographics (download here: https://osf.io/rjgwh/) - general_data.csv: contains responses to a pre-study survey (download here: https://osf.io/37uca/) After merging and cleaning the three files above, psacr001_datacleaning.Rmd exports a cleaned data file used for all analyses. This data file is stored in the data folder under two different formats: - psacr001_data_clean.rds: for those who want to analyze data in R - psacr001_data_clean.csv: for those who want to analyze data outside of R **psacr001_analyses.Rmd** psacr001_analyses.Rmd opens psacr001_data_clean.rds and runs all the analyses reported in the main text of the manuscript. **psacr001_multiverse.Rmd** psacr001_multiverse.Rmd opens psacr001_data_clean.rds and runs the multiverse analyses detailed in the supplemental information of the manuscript. In the process, it creates several output files containing the multiverse analysis results, which should be stored in a user created folder titled "multiverse.output". Copies of the output are located in the Multiverse Output folder, which is located in the Other folder on this OSF page. - anx.results.rds: results from the multiverse analysis examining the effects of framing on anxiety - behint.c.results.rds: results from the multiverse analysis examining the effects of framing on a continuous measure of behavioral intensions - behint.d.results.rds :results from the multiverse analysis examining the effects of framing on a dichotomous measure of behavioral intensions - infoseeking.rds: results from the multiverse analysis examining the effects of framing on information seeking - policy.results.rds: results from the multiverse analysis examining the effects of framing on policy support Materials ----------------- The Materials folder contains two sets of materials: - PSACR001_MainStudyMaterials_Copy.docs: the materials for the main study. - PSACCR_GeneralQuestionanire_Copy.docx: the materials used in the pre-study survey. Pre-registration ---------------- The PSAC-CR pre-registration can be found in the Registration tab of this OSF page. Alternatively, users can find a copy of this pre-registration in the Pre-Registration folder. Power analysis ---------------- This folder contains the code and output from the PSACR001 power simulations. The power simulations were executed on the [Center for High-Throughput Computing](https://chtc.cs.wisc.edu/)'s high-throughput cluster, which we needed to use due to the high computational requirements of the power analysis. Due to the reliance on a high-throughput cluster, **our workflow is not reproducible on a single local machine**. If you have access to the CHTC cluster (or, hypothetically, any cluster that uses the Condor scheduling language), you can reproduce our results using the following steps: 1. The user logs into a submit node and uses the function `condor_submit` (which is built in to the submit node) to submit `psacr001_power.sub` to the job scheduler 2. `psacr001_power.sub` tells the scheduler to queue one job for each line of `conditions.csv`, which was created prior to logging into the submit node using `create_conditions.R`. Each job runs with a different set of simulation parameters 3. The set of simulation parameters, `runR.sh`, `psacr001_power.R`, and a copy of a Linux R installation are all sent to to separate machines in the CHTC compute cluster 4. The separate machines use `runR.sh` with the simulation parameters they are sent. `runR.sh` execuites `psacr001_power.R` using R. R saves out the results of the simulation to a .csv file and sends them back to the submit node 5. The results are downloaded from the submit node and aggregated using `check_sims.R`. This file creates a .csv and figure of the simulation results. A nicely-formatted version of the results can be found in `PSACR001 power simulation.xlsx`