Approach to estimating other parameters

In the previous section, we described standard errors and confidence intervals for sample means. Population means are the most commonly estimated parameters, but there are many applications where other parameters are of greater interest. The approach is similar.

Point estimate
To estimate a population parameter, the corresponding sample statistics is used.
Error distribution
The estimate is unlikely to be equal to the population parameter. The estimation error has a distribution.
Standard error and bias
A good estimator usually has an error distribution whose mean is zero (unbiased). The standard deviation of the error distribution is called the standard error. The standard error may depend on unknown parameters.
From data:
We can usually find a numerical value for the standard error (or an approximation to it). The standard error tells you how far the estimate is likely to be from the population parameter.