Skip to content
Q28Section I (Objective (Q1–Q80))28 of 80

Statistical Inference — sufficiency in two uniform families

Consider the following statements: I. If Xi∼U(0,θ)X_i \sim U(0, \theta), θ∈(0,∞)\theta \in (0, \infty), then max⁡(X1,X2,X3,…,Xn)\max(X_1, X_2, X_3, …, X_n) is sufficient for θ\theta. II. If Xi∼fθ(x)={1,θ−0⋅5<x<θ+0⋅5 0,otherwiseX_i \sim f_\theta(x) = \begin{cases} 1, & \theta-0\cdot 5 < x < \theta+0\cdot 5 \ 0, & \text{otherwise} \end{cases}, then (min⁡Xi,max⁡Xi)(\min X_i, \max X_i) is jointly sufficient for parameter θ\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.