This preview shows pages 1–6. 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 DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Testing Hypothesis About Coe cients Con dence Intervals Regression on a Binary Variable Heteroskedasticity and Homoskedasticity GaussMarkov Theorem Introduction to Econometrics Chapter 5: Regression with a Single Regressor: Hypothesis Testing Geo rey Williams gwilliams@econ.rutgers.edu October 11, 2010 Geo rey Williams gwilliams@econ.rutgers.edu Introduction to Econometrics Chapter 5: Regression with a Testing Hypothesis About Coe cients Con dence Intervals Regression on a Binary Variable Heteroskedasticity and Homoskedasticity GaussMarkov Theorem How strong is that pattern? Just because we can t a line doesn't necessarily mean the line MEANS SOMETHING. In the previous chapter we discussed t . An even more important idea is the signi cance of the regression coe cient(s). In the case of the school districts data, we found a coe cient of 2 : 28  reducing average class size in a school district by one student was related to a 2.28 point increase in test scores. In theory the school administrators could use that information, reduce class sizes and improve results. But that would cost a huge amount of money  before you do that, you want to be sure. Geo rey Williams gwilliams@econ.rutgers.edu Introduction to Econometrics Chapter 5: Regression with a Testing Hypothesis About Coe cients Con dence Intervals Regression on a Binary Variable Heteroskedasticity and Homoskedasticity GaussMarkov Theorem How strong is that pattern? Just because we can t a line doesn't necessarily mean the line MEANS SOMETHING. In the previous chapter we discussed t . An even more important idea is the signi cance of the regression coe cient(s). In the case of the school districts data, we found a coe cient of 2 : 28  reducing average class size in a school district by one student was related to a 2.28 point increase in test scores. In theory the school administrators could use that information, reduce class sizes and improve results. But that would cost a huge amount of money  before you do that, you want to be sure. Geo rey Williams gwilliams@econ.rutgers.edu Introduction to Econometrics Chapter 5: Regression with a Testing Hypothesis About Coe cients Con dence Intervals Regression on a Binary Variable Heteroskedasticity and Homoskedasticity GaussMarkov Theorem Hypothesis Testing! Again! Just as we did in chapter 3, we can formulate a 2sided or 1sided hypothesis. Twosided version would be: H : 1 = 1 ; vs. H 1 : 1 6 = 1 ; A onesided version would be: H : 1 = 1 ; vs. H 1 : 1 &lt; 1 ; or H : 1 = 1 ; vs. H 1 : 1 &gt; 1 ; Geo rey Williams gwilliams@econ.rutgers.edu Introduction to Econometrics Chapter 5: Regression with a Testing Hypothesis About Coe cients Con dence Intervals Regression on a Binary Variable Heteroskedasticity and Homoskedasticity GaussMarkov Theorem Hypothesis Testing! Again!...
View
Full
Document
This note was uploaded on 01/22/2012 for the course ECONOMICS 220:322 taught by Professor Otusbo during the Fall '10 term at Rutgers.
 Fall '10
 Otusbo
 Econometrics

Click to edit the document details