pam2100 stock_watson_ch5

# pam2100 stock_watson_ch5 - 5-1 Also we will cover some...

This preview shows pages 1–4. Sign up to view the full content.

Stock and Watson Chapter 5: Regression with a Single Regressor what do the distributions of 1 ˆ β and 0 ˆ look like? if the sample is sufficiently large (n=100), then: 1 ˆ ~ N (β 1 , standard deviation given by formula on page 133 SW, 587 MM) 0 ˆ ~ N (β 0 , standard deviation given by formula on page 133 SW, 587 MM) we don’t know the value of these standard deviations (depend on mean and variance of X )- the estimated standard deviations of 1 ˆ and 0 ˆ are the standard errors of 1 ˆ and 0 ˆ : if the sample is sufficiently large ( n =100), then: 1 ˆ ~ N (β 1 , standard error given by formula on page 151 SW, 587 MM) 0 ˆ ~ N (β 0 , standard error given by formula on page 180 SW, 587 MM) t = ) ˆ ( _ ˆ 1 1 SE value ed hypothesiz - 1) if the sample is sufficiently large, then 1 ˆ ~ N 2) the t -statistic is calculated assuming that the null is true H 0 : β 1 = 0 vs. H 1 : β 1 0 if the null is true, then the mean ( 1 ˆ ) = 0 3) SE ( 1 ˆ ) is the estimated standard deviation of 1 ˆ 1 ˆ ~ N t -statistic is (approximately) a standardized variable→ (variable – mean) / (estimated sd) t -statistic has (approximately) standard normal distribution: t ~ N (0,1) → can use standard normal table to calculate p-value

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
if the sample is sufficiently large→ t ~ N (0,1) also legitimate to write: if the sample is sufficiently large→ t ~ t n -2 since t n -2 approaches the standard normal distribution as n gets large (return to t distributions in Section 5.6) Stock and Watson use: if the sample is sufficiently large→ t ~ N (0,1) we will always use t ~ N (0,1) unless explicitly told otherwise (Section 5.6) when will t statistic have exactly a t n -2 distribution? (discuss in Section 5.6) Example 1 estScore T ˆ = 700 - 2.32 × STR n = 420 (10.4) (1.12) Calculate the t -statistic for the test of H 0 : β 1 = 0 against the two-sided alternative H a : β 1 0? Calculate the p-value using the standard normal table. Do you reject H 0 at the 5% level? Example 2 estScore T ˆ = 700 - 2.32 × STR n = 420 (10.4) (1.12) Calculate the t -statistic for the test of H 0 : β 1 = 0 against the one-sided alternative H a : β 1 < 0? Calculate the p-value using the standard normal table. Do you reject H 0 at the 5% level? 2
Example 3 estScore T ˆ = 698.9 - 2.28 × STR n = 420 (10.5) (0.52) Construct a 95% confidence interval for β 1 . Use a z * value that refers to the standard normal distribution. Confidence intervals for predicted effects of changing

This preview has intentionally blurred sections. Sign up to view the full version.

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

## This note was uploaded on 10/17/2010 for the course PAM 2100 taught by Professor Lewis during the Spring '10 term at Adelphi.

### Page1 / 11

pam2100 stock_watson_ch5 - 5-1 Also we will cover some...

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online