Skip to content
Q6Section I (Objective (Q1–Q80))6 of 80

Statistical Inference — variance of the sample variance

Let X1,X2,X3,…,XnX_1, X_2, X_3, …, X_n be a random sample from N(μ,σ2)N(\mu, \sigma^2). Consider the statistic s2=1n−1∑i=1n(Xi−X‾)2s^2 = \frac{1}{n-1}\sum_{i=1}^{n}(X_i-\overline{X})^2. What is V(s2)V(s^2) equal to?

Options

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