There are many tasks in behavioral economics where individuals repeatedly interact with and learn about their
environment. When researchers analyze these sorts of tasks, they often formulate bayesian learning models
to try and fit subject behavior. An individual's 'prior', or prior probability distribution, is essentially
a representation of what the learner knows before ever receiving feedback about their decisions.
Some models can be greatly improved by explicitly asking individuals for their prior. The distribution
builder tool allows individuals to simply draw their prior distribution, and outputs a single array for
researchers to incorporate in their analysis. All you have to do is click and start drawing. The code for
this project can be easily modified to accommodate any number of bars and canvas size.
Below is an example of the how the tool works.
There are ten poker players that you can choose to compete against, each is skilled but you do not know
how much better or worse they are than you. For each player (1-10) indicate how likely you think you are
to win against them (0-100).