The basis of confidence intervals holding any usefulness is pretty shaky. See http://uncertainty.stat.cmu.edu/ http://uncertainty.stat.cmu.edu/ (it also has a c

The basis of confidence intervals holding any usefulness is pretty shaky. See http://uncertainty.stat.cmu.edu/ http://uncertainty.stat.cmu.edu/ (it also has a chapter on old ideas such as this).