Blavatskyy (2011) Calculator

This calculator implements Blavatskyy's (2011) model of probabilistic choice. The example R code implements the model for a toy three-way choice. Pressing "run" submits the code to R and generates predictions, or you can edit the code for your own choices. Predictions appear at the bottom of the output page, and give the probability of selecting each gamble in each choice.

Full source code for the model is available, with a worked example applying the model to the choices from Kahneman and Tversky (1979).

rm(list=ls()) # Load the Blavatskyy (2011) library source("Blavatskyy_2011.R") # Set up three gambles G1 <- gamble(p=c(.2, .3, .5), x=c(20, 40, 60)) G2 <- gamble(p=c(.1, .25, .65), x=c(10, 30, 70)) G3 <- gamble(p=c(.5, .1, .4), x=c(0, 30, 100)) # Make one choice from the gambles (choice <- list(G1=G1, G2=G2, G3=G3)) # Run the Blavatskyy (2011) model to get the probability of selecting each # gamble Blavatskyy(choice, gamma=0.5, lambda=0.5) # The default model assumes a power law utility function with exponent gamma # and the form phi(u) = exp(-lambda * u) -1 for the transformation of # utilities (see Blavatskyy, 2011, p. 4). Alter the functions in # P.specific() for different assumptions

You can attach data from an external file, though you do not need to do so for the default example. The file should be a text file, with white spaces delimiting the fields, and column headings beginning with a letter and containing only letters, digits, and '.'. The data file is stored as a data.frame in X, and the variables within X are attached. For example, for the data file below, X$a=c(1, 4, 7), and a=c(1,4,7).

a b c
1 2 3
4 5 6
7 8 9

Select a local file to submit: