slides_Ch2_W[1]

# 2 an example q1a to see that the unbiased estimator y

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: le) fA ) ( E [ Y j A ] E [Y jB ]) fA ) is known. Econometrics 66 / 93 Failure of SLR.2: An Example (Q.1) Use what we showed to “guess” an unbiased estimator: E [ y ] = E [ Y ] + ( sA fA ) ( E [ Y j A ] E [Y jB ]) Observe that by C.1 (sA fA ) is known. Observe that by C.2 we have unbiased estimators of E [Y jA] and E [Y jB ]: E [y A ] = E [Y jA] E [y B ] = E [Y jB ] Melissa Tartari (Yale) Econometrics 66 / 93 Failure of SLR.2: An Example (Q.1) Use what we showed to “guess” an unbiased estimator: E [ y ] = E [ Y ] + ( sA fA ) ( E [ Y j A ] E [Y jB ]) Observe that by C.1 (sA fA ) is known. Observe that by C.2 we have unbiased estimators of E [Y jA] and E [Y jB ]: E [y A ] = E [Y jA] E [y B ] = E [Y jB ] Thus, C.1 and C.2 imply that the following estimator, called y W , is unbiased for E [Y ]: yW Melissa Tartari (Yale) y ( sA fA ) ( y A Econometrics yB ) . 66 / 93 Failure of SLR.2: An Example (Q.1) We will next answer two questions: Melissa Tartari (Yale) Econometrics 67 / 93 Failure of SLR.2: An Example (Q.1) We will next answer two questions: Q.1.a Does the unbiased estimator y W make intuitive sense? Melissa Tartari (Yale) Econometrics 67 / 93 Failure of SLR.2: An Example (Q.1) We will next answer two questions: Q.1.a Does the unbiased estimator y W make intuitive sense? Q.1.b What does this have to do with regression analysis in the SLRM? Melissa Tartari (Yale) Econometrics 67 / 93 Failure of SLR.2: An Example (Q.1.a) To see that the unbiased estimator y W is intuitively sensible it is useful to do some algebra ... yW y = ... 1 = N = Melissa Tartari (Yale) 1 N ( sA N fA ) ( y A fA N ∑ sA yi + N ∑ 1 fA yi sA ∑ ωi yi fj ( i ) sj ( i ) i =1 N i =1 Econometrics 1 yB ) 1 i =1 with ω i = 68 / 93 Failure of SLR.2: An Example (Q.1.a) To see that the unbiased estimator y W is intuitively sensible it is useful to do some algebra ... yW y = ... 1 = N = 1 N ( sA N fA ) ( y A fA N ∑ sA yi + N ∑ 1 fA yi sA ∑ ωi yi fj ( i ) sj ( i ) i =1 N i =1 1 yB ) 1 i =1 with ω i = We see that y W is a weighted sample aver...
View Full Document

Ask a homework question - tutors are online