#46 Wrong variance in JAGS man for logistic distn

[Anonymous]

There appears to be an error in the JAGS 4.3.0 manual (28 June 2017). On page 46 the variance of the logisitc distribution is stated to be (pi^2)/(3 x tau). This variance should be (pi^2)/(3 x tau^2). Under this parameterization it seems a little confustion to me to call tau the precision parameter, tau is proportional to the sqrt of the what I would call the precision parameter.

[Martyn Plummer]

hanks for this. I made similar mistakes in the documentation for other distribution (notably the double exponential). I suppose that subconsciously I was expecting more consistency in the parameterization for location-scale distributions in the BUGS language.

I have corrected the variance formula and now refer to the parameter tau as the "inverse scale"