Skip to content
Q1Section I (Objective (Q1–Q80))1 of 80

Statistical Inference — concentration versus mean square error

Consider the following statements: I. The concentration of distribution of statistic T(x)T(x) around its true value θ\theta is high when the mean square error is low but the converse is not true. II. There exists no estimator for which the mean square error is least for all θ∈Θ\theta \in \Theta. Which of the statements given above is/are correct?

Options

Pick an option to check your answer. In a mock test a wrong answer here costs 0.83 mark.