Change the range not the language on confidence intervals

I agree with the last sentence and with the better label of uncertainty interval. But as a definitional device this is not enough help to a researcher in my opinion. This fits in with my old bias that frequentist methods are only simple and helpful if you obfuscate them. As a side note, the great Bayesian biostatistician Don Berry has said that he became a Bayesian over frustration at teaching confidence intervals to smart biostatistics graduate students.