On this page
Identify two different means
A fictional sample produces observed mean 7.2. The expectation of a future sample mean under the same population model is mu. Unless mu is independently known to be 7.2, the observed estimate does not define the actual center of repeated sampling.
Check an enumerated example
The population values 2,4,6,8 have mean 5. One size-two sample consisting of 2 and 4 has mean 3. The full distribution over all six equally likely samples is still centered at 5. The chosen sample has not moved the population center to 3.
Label an estimated or resampling model
An analyst can fit a model or construct an empirical resampling distribution centered near the observed estimate. That is an estimated model, not direct knowledge of the underlying sampling distribution. Keep the two claims separate.
Use honest parameter language
Write that x-bar estimates mu and that s/√n estimates a standard error under the appropriate independent model. Avoid presenting an observed sample mean as the known population mean while simultaneously discussing uncertainty about that mean.