Lecture 2

Question 1

On a linear regression model y=a+bx+e what does b measure?

Answer 1

B shows the increase in y when x increases by 1 unit

Question 2

What does B measure in a log transformation model: Y=a+blog(x) +e

Answer 2

This measures: If x changes by 1% the percentage of Y changes by B%

Question 3

What does B measure in the model Log(y)=a+bx+e

Answer 3

This shows that when x changes by 1 unit 100B in the % change in Y

Question 4

What does B measure in log(y)=a+Blog(b)+e

Answer 4

When x increases by a percentage point Y increases by B% or by B/100

Question 5

What rule can we apply if changes are small?

Answer 5

ln(1 + growth rate) = growth rate if the growth rate is less than 0.1

Question 6

What are the 4 steps in conducting a hypothesis test?

Answer 6

1. Specify the null hypothesis Ho:B=Bo
2. Specify the alternative hypothesis: H1:B doesn’t equal Bo
3. Choose a significance level and compute the test statistic
4. Compare the test statistic to the critical value
5. If the Test statistic is within the critical region reject the null hypothesis

Question 7

How do we know if a result is unbiased?

Answer 7

E(b|x)=B is unbiased if the average of the OLS estimator = the target parameter B

Question 8

Be able to show E(b|x)B and also estimate the variance

Answer 8

See notes

Question 9

Be able to conduct a measure of goodness of fit test

Answer 9

See notes