This preview shows pages 1–4. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Regression with a Single Regressor: Hypothesis Tests and Confidence Intervals (SW Chapter 5) Object of interest: 1 in, Y i = + 1 X i + u i , i = 1,, n 1 = Y / X , for an autonomous change in X ( causal effect ) The Least Squares Assumptions: 1. E ( u  X = x ) = 0. 2. ( X i ,Y i ), i =1,, n , are i.i.d. 3. Large outliers are rare ( E ( X 4 ) < , E ( Y 4 ) < . General setup Null hypothesis and twosided alternative: H : 1 = 1,0 vs. H 1 : 1 1,0 where 1,0 is the hypothesized value under the null. Null hypothesis and onesided alternative: H : 1 = 1,0 vs. H 1 : 1 < 1,0 General approach : construct tstatistic, and compute pvalue (or compare to N (0,1) critical value) In general: t = estimator  hypothesized value standard error of the estimator where the SE of the estimator is the square root of an estimator of the variance of the estimator. For testing the mean of Y : t = ,0 / Y Y Y s n  For testing 1 , t = 1 1,0 1 ( ) SE  , where SE ( 1 ) = the square root of an estimator of the variance of the sampling distribution of 1 Formula for SE( 1 ) Recall the expression for the variance of 1 (large n ): var( 1 ) = 2 2 var[( ) ] ( ) i x i X X u n  = 2 4 v X n , where v i = ( X i X ) u i . SE ( 1 ) = 1 2 = the standard error of 1 Summary: To test H : 1 = 1,0 v. H 1 : 1 1,0 , Construct the tstatistic t = 1 1,0 1 ( ) SE  = 1 1 1,0 2  Reject at 5% significance level if t > 1.96 The pvalue is p = Pr[ t  >  t act ] = probability in tails of normal outside  t act ; you reject at the 5% significance level if the pvalue is < 5%. This procedure relies on the large n approximation; typically n = 50 is large enough for the approximation to be excellent. Example: Test Scores and STR , California data Estimated regression line: test score = 698.9 2.28 STR Regression software reports the standard errors: SE ( ) = 10.4 SE ( 1 ) = 0.52 tstatistic testing 1,0 = 0 = 1 1,0 1 ( ) SE  = 2.28 0 0.52 = 4.38 The 1 % 2sided significance level is 2.58, so we reject the null at the 1% significance level. Alternatively, we can compute the pvalue The pvalue based on the large n standard normal approximation to the tstatistic is 0.00001 (10 5 ) Because the tstatistic for 1 is N(0,1) in large samples, construction of a 95% confidence for 1 is just like the case of the sample mean: 95% confidence interval for 1 = { 1 1.96 SE ( 1 )} Confidence interval example : Test Scores and STR Estimated regression line: testscore= 698.9 2.28 STR SE ( ) = 10.4 SE ( 1 ) = 0.52 95% confidence interval for 1 : { 1 1.96 SE ( 1 )} = {2.28 1.96 0.52} = (3.30, 1.26) A concise (and conventional) way to report regressions:...
View Full
Document
 Fall '09
 Li
 Interest

Click to edit the document details