We say things like "The $(1-\alpha)100$% PI coverage measures the proportion of the time that PIs at that nominal level included the true value, which provides information about whether a forecast has accurately characterized its uncertainty about future observations. Achieving approximately nominal ($(1-\alpha)100$%) coverage indicates a well-calibrated forecast."
I can imagine two groups of readers:
One group is people who have taken and understood a statistics course. Those people already understand what interval coverage rates are, so these sentences are not targeted to those readers.
The other group is people who have not taken and understood a statistics course. Those readers will wonder what alpha is and why we are talking about it.
So maybe we can rework these sentences to avoid the (1 - alpha) * 100% formalism?
We say things like "The $(1-\alpha)100$% PI coverage measures the proportion of the time that PIs at that nominal level included the true value, which provides information about whether a forecast has accurately characterized its uncertainty about future observations. Achieving approximately nominal ($(1-\alpha)100$%) coverage indicates a well-calibrated forecast."
I can imagine two groups of readers:
So maybe we can rework these sentences to avoid the (1 - alpha) * 100% formalism?