econometrics 1

what does z need to satisfy

instrument relevance - cov(z,x)=/=0

instrument exogeneity - cov(z,u)=0

why you can use 2sls with valid z

p2

consistency of 2sls

p3

properties of 2sls (one regressor)

ie change in z changes x how

write first stage and reduced form

p4

general iv case

p5

general properties of 2SLS

consistent, asymptotically normal = construct T and F tests

p7

2SLS in matrix

p9

proof 2SLS is NOT unbiased

p11

properties of 2SLS with poor iv

ie when is OLS better than 2SLS

p12

check instrument relevance of 2SLS

p13

check instrument exogeneity & 3 ways to check

p14

-structure of econ model

-quasi experiments - ensuring randomisation was properly done or natural events may generate randomness

-overidentification test

overidentification test

p15

find B1 hat in OLS with 1 regressor

1-2 page 4

write OLS in matrix form

write B hat in matrix form

1-2 p5

proof ui hat is uncorrelated to regressors in OLS

1-2 p9

proof of TSS = ESS + SSR

1-2 p10

proof of corrolaries

““

moment conditions for OLS 1 regerssor

1-2 p12

proof that OLS is unbiased

3-4 p3

variance of ols

3-4 p4

homoskedasticity:
when is OLS homoskedastic,

exogeneity & homoskedasticity,

variance estimation under homoskedasticity

3-4 p5

when does homoskedasticiity not hold

different variances of errors

errors are correlated

OLS matrix proof for consistency

3-4 p9

proof of asymptotic normality

““

F test - multiple parameters

5-6 p4

test stat for F test

5-6 p7

OV bias

5-6 p9

simultaneity (causal effect by Y on X)

5-6 p10

measurement error in outcome variable Y

5-6 p11

measurement error in regressors

consistency depends on cov between error residual and x1

5-6 p12