Specification and Data Issues

Question 1

3 main topics

Answer 1

1) test for misspecification
2) dealing with outliers
3) bootstrap method for calc SE

Question 2

Testing for misspecification

Answer 2

• omitted functions of explanatory variables
• log (Wage) = B0+B1Educ+B2exper+u
  omitting exper2 result in biased estimator because exper ^2 can be correlated with education
• if omitted, it will not properly describe / maximize the relationship of X to Y

Question 3

Test for misspecification

Answer 3

1) Ramsey REST Test

2) David Mackinnon Test

Question 4

Ramsey REST Test

Answer 4

• reg specification error test
• to see if non-linear functions of Xi are significant
• determine how many function of fitted values to include in expanded reg (^2/^3)

Question 5

hypothesis

Answer 5

H0 = S1 = S2 = 0 (no misspecification)
H1 = S1 = S2 =/= 0
(misspecification), there’s omitted relationship

Question 6

positives of test

Answer 6

preserves degrees of freedom

Question 7

negatives of test

Answer 7

does not indicate specific source of misspecification

Question 8

David Mackinnon Test
(nonnested alternative)

Answer 8

• decide whether an IND Var should be in level or log form

Question 9

test 1

Answer 9

only tests y hat

Question 10

test 2

Answer 10

tests log and y hat

Question 11

Outliers

Answer 11

• observation that when removed from regression, results substantial change in OLS estimator

Question 12

bootstrap method for estimating SE

Answer 12

• about what happens when statistical inference is unavailable or unreliable

Question 13

what causes hyp testing to be invalid

Answer 13

• errors are not normally distributed

Question 14

Monte Carlo Approach

Answer 14

replicate DGP thus, derive parameters of the sampling distribution

Question 15

MCA steps

Answer 15

1) Run OLS on orig data, treat est. as true parameters value

2) treat the values of explanatory variables as “Fixed in repeated samples”

3) generate value on DEP VAR based on error from random number generator

4) estimate new parameters

5) repeat 1000+ times

6) calculate stand dev of parameters

Question 16

MCA drawback

Answer 16

• generating new errors for “repeated samples” requires specifying the distribution from which those errors came

Contradicts: we don’t know the distribution is the problem in the 1st place

Question 17

Bootstrap Approach

Answer 17

• special case of Monte Carlo where errors is not pre-determined
• instead of random no. generator, use random sampling of existing residuals form orig sample