Economics
Saved & recent

Sampling uncertainty and confidence intervals

Estimates vary across samples. A confidence interval comes from a procedure with a stated long-run coverage rate under its assumptions; it does not resolve design bias.

Read first
Sampling and measurement
Related
Hypothesis tests and p-values, Causal identification
Further reading
CORE Econ: Measuring management practices — Working in Excel
true mean 10 1 2 3 4 5 6 7 8 misses mean 6 8 10 12 14 covers true mean · 7 of 8
Hypothetical intervals in outcome units, illustrating a nominal 95% procedure; eight intervals need not achieve exactly 95% realized coverage. The 95% describes long-run coverage under the model, not a 95% probability that the fixed mean lies in a particular realized interval. Sampling intervals do not remove design bias.

View in the full Economics collection