ABSTRACT
Given a sample from a family of distributions {P θ: θ ∈ Ω} indexed by a parameter θ, we would like to find a statistic whose value will be “close” to q(θ). Chapter 8 was devoted to the construction of estimators. Now we would like to measure the performance of competing estimators. For example, the “mean squared error” of a real-valued estimator T of a real parameter q(θ) based on a sample X = (X1 ..., Xn) is defined to be
