How to set priors without standard deviations

In the table 1 in Kam et al. (2009), “Uncovering the Hit List for Small Inflation Targeters: A Bayesian Structural Analysis”,
prior parameter densities are described but there are only prior means and 2.5% and 97.5% confidence interval. That is, there are no standard deviations.

Is there any way to recover prior standard deviation or to use dynare with such information?

Also, in their paper, parameters in loss functions are also estimated.
However, it is totally nonlinear because the loss function is quadratic.
I would like to know how to estimate the policy preference parameters in the loss function using dynare.

If you have Matlab’s Statistics toolbox, you can use the codes at github.com/JohannesPfeifer/prior_from_quantiles

Unfortunately, it is not easily possible to estimate models under optimal policy with Dynare. This is on our to-do list: github.com/DynareTeam/dynare/issues/1173