Parameterizing varying-intercept, varying-slope models: the folder nes contains data from the National Election Study surveys. Set up a model for party identification (as a continuous outcome, as in...

Parameterizing varying-intercept, varying-slope models: the folder nes contains data from the National Election Study surveys. Set up a model for party identification (as a continuous outcome, as in Section 4.7), given the predictors shown in Figure 4.6 on page 74, and also allowing the intercept and the coefficient for ideology to vary by state. You will fit various versions of this model (using data from the year 2000) in Bugs.

(a) Fit the model with no correlation between the intercepts and the slopes (the coefficients for ideology).

(b) Fit the model with a uniform prior distribution on the correlation between the intercepts and slopes.

(c) Fit the scaled inverse-Wishart model.

(d) Compare the inferences from the three models.