Statisticians routinely construct interval estimates by making their point estimate be the interval center and creating a range of other possible values, known as the error bound, below and above the center.
The error bound is a half-width of an interval estimate, equal to the difference between the point estimate on the one hand and either the lower or upper limit of the interval on the other hand.
The unknown parameter is presumed to lie at the center of the interval that the point estimate and the error bound create.
The error bound takes only sampling error into account, ignoring all other potential sources of error, such as bias in questions, selection bias, or miscalculations.
Your Opinion
Quick Reply
