If the population standard deviation is known, rather than estimated, will a 95% confidence interval be wider in a situation where the z-distribution is used (i.e., σ known) or when a t-distribution is used (i.e., σ unknown)??
-
Maria
For any given confidence interval, a t-interval will always be ≥ compared to the width of the z-interval.
As the degrees of freedom approaches infinity, the widths will be equal.
Hope that helps
For any given confidence interval, a t-interval will always be ≥ compared to the width of the z-interval.
As the degrees of freedom approaches infinity, the widths will be equal.
Hope that helps