Inferential Statisticshard
0:00.0

In the context of the Wald test for H0:θ=θ0H_0: \theta = \theta_0, suppose the estimator θ^n\hat{\theta}_n follows n(θ^nθ)dN(0,V(θ))\sqrt{n}(\hat{\theta}_n - \theta) \xrightarrow{d} N(0, V(\theta)). If the null hypothesis is true, what is the asymptotic behavior of the Wald statistic Wn=n(θ^nθ0)T[V(θ^n)]1(θ^nθ0)W_n = n(\hat{\theta}_n - \theta_0)^T [V(\hat{\theta}_n)]^{-1}(\hat{\theta}_n - \theta_0)?