Replication data for: Offline volunteering during COVID-19: A survey experiment with prior and prospective blood donors
Description
Readme for Offline volunteering during COVID-19: A survey experiment with prior and prospective blood donors, by Stefanie Ehmann, Michael Haylock, and Anne Heynold. Uploaded data includes only intentions to donate blood and an experimental treatment (information about blood shortages during COVID-19 and hygiene measures during COVID-19 at donation centers).
Files
Steps to reproduce
Readme for Offline volunteering during COVID-19: A survey experiment with prior and prospective blood donors, by Stefanie Ehmann, Michael Haylock, and Anne Heynold. Uploaded data includes only intentions to donate blood and an experimental treatment (information about blood shortages during COVID-19 and hygiene measures during COVID-19 at donation centers). No personal data about previous donations or blood groups are included. All analyses were performed in R studio on a windows computer. The survey was run in SoSciSurvey, which also generated a data import code for R. To estimate main results (Results for Column 1 Table 3, histogram with all subjects in Fig 1, Table 4 Column 1), load Replication_data.Rda into RStudio (we use Version 4.3), and then run the code Replication_results.R This will generate the main treatment effects at intensive and extensive margin, as well as the histogram of all respondents by treatment. In our analyses, we ran: Raw data: raw_data.csv (not available publicly due to sensitive information) Master.r runs the following files: 1) Import data using: Data_import.r 2) Clean data using: Data_clean.r 3) Table_1_2_3_A2_A3 generates results at extensive margin, summary stats, and self-reported past donations. 4) Fig_1_2_A2_Tables_A4_A5_A6_A7 generates results from the intensive margin analyses 5) Replication data: Data_replication generates a reduced data file with blood donation intentions at the extensive margin outcome with (futuredonor: TRUE="yes", FALSE="no") and intensive margin (speed: "Next week" = 1, "Next month" = 2, "After lockdown" = 3, "Not donate" = 4), as well as the experimental treatment (“Control" = 1, "Shortage" = 2, "Hygiene" = 3). Aftermininterval descibes whether the respondent said they had just donated and are waiting (True or false in excel, labeled categories in R in the R data file) . If there are any questions or additional results that are requested, please contact the authors Michael Haylock (michael.haylock@uni-due.de) and provide him with an R script that fits the data and variables described in the R code, so that these can be run by the authors.