Study title

The collection contains four sets of data files and related questionnaires and experimental scripts, preceded by three files that contextualise the data. The file "project overview.pdf" describes the project aims, how the data was collected, main findings and outputs such as publications and presentations. The file "Consent procedure.docx" describes how we elicited informed consent from research participants for collecting, archiving and sharing the data.The file "readme.txt" contains an index of all data files. Note that each of the four data sets is in a zipped folder. Each of the folders contains a file with the instructions and/or questionnaires for participants, as well as all relevant data files. All data files are self-explanatory, i.e. have variable and value labels. 1. The files "Probability weighting and subjective probabilities (data and scripts)" contain experimental data for common consequence ladders (probability weighting) and tempting fate and fatalism experiments (subjective probabilities). The scripts used for the three experiments are provided and serve as a guide to the data. 2. The files "Nature's frames experiment data and script" contain the data for a test of reference-dependent utility theory. The script used in the experiment is provided and serves as a guide to the data. 3. The files "Risk pooling experiment data and script" contain the data for a risk sharing experiment (with a focus on implications for risky investment behaviour). The script used in the experiment is provided and serves as a guide to the data. 4. The files "Household survey (data and questionnaire" contain the data for the household survey that accompanied the experiments; the questionnaire is provided too. The variable "token" is the unique ID that may be used for linking the household survey data and the experimental data.Farmers in developing countries operate in extraordinarily hazardous environments. Policies that aim to increase agricultural productivity in such...