slides_Ch2_W[1]

# melissa tartari yale econometrics 89 93 the variance

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: or that is not unbiased. The unbiased estimator makes a degrees-of-freedom adjustment ˆ and we denote it by σ2 : SSR ˆ σ2 = . (2.61) n2 Melissa Tartari (Yale) Econometrics 85 / 93 The Variance of the OLS Estimator - Theorem 2.3 Theorem 2.3: under assumptions SLR.1-SLR.5 ˆ E σ 2 jx = σ 2 . Melissa Tartari (Yale) Econometrics 86 / 93 The Variance of the OLS Estimator - Theorem 2.3 Theorem 2.3: under assumptions SLR.1-SLR.5 ˆ E σ 2 jx = σ 2 . ˆ ˆ ˆ Thus, if we plug σ2 into Var β1 jx and Var βo jx we have unbiased \ \ ˆ ˆ estimators, which we denote Var β1 jx and Var βo jx . (Incidentally: How can we say that they are unbiased? Because they are linear functions of unbiased estimators and the expected value is a linear operator). Melissa Tartari (Yale) Econometrics 86 / 93 The Variance of the OLS Estimator - Theorem 2.3 - proof ˆ We are interested in E σ2 jx and we know that (n ˆ 2) σ 2 = n ˆ ∑ ui2 , ) E (n i =1 n ˆ ˆ 2) σ2 = E [ ∑ ui2 jx ] i =1 therefore we need to work out the RHS. Melissa Tartari (Yale) Econometrics 87 / 93 The Variance of the OLS Estimator - Theorem 2.3 - proof ˆ We are interested in E σ2 jx and we know that (n ˆ 2) σ 2 = n ˆ ∑ ui2 , ) E (n i =1 n ˆ ˆ 2) σ2 = E [ ∑ ui2 jx ] i =1 therefore we need to work out the RHS. Consider (2.59) and (a) its averaged-form, (b) di¤erence from its averaged form: ui ˆ = ui 0=u ui ˆ Melissa Tartari (Yale) = (ui ˆ βo ˆ β o u) βo βo ˆ β Econometrics ˆ β1 ˆ β 1 1 β1 xi β1 x β1 (xi x) (2.59) (a) (b) 87 / 93 The Variance of the OLS Estimator - Theorem 2.3 - proof Next we (c) square both sides and then (d) sum both sides: ˆ u )2 + β1 ˆ 2 (ui u ) β ui2 = (ui ˆ 1 n ˆ ∑ ui2 i =1 n = ∑ (ui β1 ˆ u )2 + β1 x )2 (xi β1 (xi β1 x) 2 (c) n ∑ (xi x )2 i =1 i =1 ˆ 2 β1 2 n β1 ∑ (ui u ) (xi x) (d) i =1 Melissa Tartari (Yale) Econometrics 88 / 93 The Variance of the OLS Estimator - Theorem 2.3 - proof We are now in the position of taking expectations of both sides and write (( ) on 2...
View Full Document

Ask a homework question - tutors are online