Choosing Beta Priors Interactively

less than 1 minute read

Choosing Beta prior distributions can be difficult when using the default parameterisation because the shape parameters are difficult to interpret. To get around this, I’ve written a simple rshiny app for teaching purposes, where the user can play with sliders to change the mean and spread of a Beta distribution and visualise the shape at the same time. Of course, the app translates the mean and spread back to alpha and beta parameters so that once a good prior has been found, it can be implemented in R.