Next: Point Estimation
Previous: Useful Sampling Distributions

An estimator is a rule (or a formula, or a statistic) that tells us how to use the data contained in a sample from some population, to estimate the value of some characteristic (i.e., parameter) of the population.

We will use the generic symbol q to denote the (unknown) value of the parameter.

There are two types of estimates (and thus two types of estimators):

• Point Estimate: A single value that is used as an estimate of the true (unknown) value of the population parameter. The corresponding estimator (i.e., formula or statistic) used to compute this value is called a point estimator. As a matter of convenience, we denote the estimator as well as the estimate (i.e., the statistic as well as its value) by . Thus, for example if the parameter of interest is the mean m , then its point estimate is denoted by , while if we the parameter of interest is the S.D. s, its point estimate is denoted by .

• Interval Estimate: An interval within which we estimate the true value of the parameter to lie. Thus the interval estimate will consist of two values - say a and b - and we estimate that the true value q lies in the interval [a,b]. The interval is more commonly referred to as a confidence interval.