Econ 281 Regression Analysis with One Variable

Econ 281 Regression Analysis with One Variable - Regression...

Info iconThis preview shows pages 1–14. Sign up to view the full content.

View Full Document Right Arrow Icon
1 Regression Analysis with One Variable Introduction to Econometrics Econ 281
Background image of page 1

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

View Full DocumentRight Arrow Icon
2 Introduction Remember our two examples Student Test Scores and Student-Teacher Ration Wages and Education Both are Relationships between two variables Dependent Variable: Test Scores, Wage Independent Variable (Regressors): Student-Teacher Ration, Education Simple Regression Analysis is all about investigating the relationship between those two variables
Background image of page 2
3 Introduction We start out by “looking” at the population relationship between the two variables Note: We are not really able to do that The population regression function: m(x i ) is a general function Example: () ( ) i i i x m x X Y E = = | ( ) i i i edu m edu Education wage E = = |
Background image of page 3

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

View Full DocumentRight Arrow Icon
4 Saturated Models One can keep the population regression function completely flexible by including dummy (binary) variables and their interactions Dummy Variables = indicator variable that is equal to one if a condition is met and otherwise equal to zero 1 0 male if individual is male D otherwise =
Background image of page 4
5 Saturated Models Interaction = the product of two or more variables Let’s assume that we only have two variables: gender and race (white, non- white) A saturated model looks like ( ) 01 2 3 |, i male gender male gender male gender EY D D D D DD ββ β =+ + +⋅
Background image of page 5

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

View Full DocumentRight Arrow Icon
6 Saturated Models Let’s see why this is a perfectly general model There are four outcomes ( ) 0 |0 , 0 i male gender EY D D β == = ( ) 01 |1 , 0 i male gender D D = + ( ) 02 , 1 i male gender D D = + ( ) 012 3 , 1 i male gender D D ββ = + + +
Background image of page 6
7 Saturated Models Why is this perfectly general? => Every possible constellation has its own effect => No need to combine effects However, we usually don’t use saturated models. Why? How many possible outcomes do you have when there are 10 possible educational outcomes 10 possible experience outcomes Gender Race Too many!! We need to simplify things!
Background image of page 7

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

View Full DocumentRight Arrow Icon
8 Linearity Assumption We will simplify things by assuming that m(x i ) is linear Note: It is important to remember that this is an assumption and that we NEVER actually observe the TRUE relationship between those variables i i X x m + = 1 0 ) ( β
Background image of page 8
9 Linearity Assumption Now we “confront” this model with data taken from the population. Would we expect the above relationship to hold perfectly? No! Data differs because other variables influence Y i Data differs because of random variation i i X Y + = 1 0 β
Background image of page 9

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

View Full DocumentRight Arrow Icon
10 Linearity Assumption Solution: We add an ERROR TERM (u i ) to the equation The error term captures the effect of other variables random deviations from Y i This gives us the linear regression model i i i u X Y + + = 1 0 β
Background image of page 10
11 The Linear Regression Model i i i u X Y + + = 1 0 β Slope Coefficient Intercept Coefficient Dependent Variable Independent (Regressor) Error Term
Background image of page 11

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

View Full DocumentRight Arrow Icon
12 (X 1 ,Y 1 ) 600 620 640 660 680 700 15 20 25 Test Scores and Student-Teacher Ratio Student-Teacher Ratio Test Scores (Unobserved) Population Regression Function =u 1 (X 2 ,Y 2 ) (X 3 ,Y 3 ) =u 2 =u 3 (X 4 ,Y 4 ) =u 4
Background image of page 12
13
Background image of page 13

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

View Full DocumentRight Arrow Icon
Image of page 14
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 09/25/2010 for the course ECON 281 taught by Professor Habermalz during the Winter '08 term at Northwestern.

Page1 / 126

Econ 281 Regression Analysis with One Variable - Regression...

This preview shows document pages 1 - 14. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online