Department of Economics
Princeton University
Economics 302
Econometrics
Syllabus, Fall 2013
(09-10-2013)
Lectures:
MW 1:30-2:50PM
Payton Hall 145
Professor:
Bo E. Honor, 209 Fisher Hall
Office Hours: Mondays 3:00-5:00 PM.
(Meetings should be scheduled via
Stata Tutorial1
Logging into Stata Remotely
Both PC and Mac users need to register their accounts to use UNIX servers and the Nobel server
in particular.
For UNIX accounts go here: http:/helpdesk.princeton.edu/kb/display.plx?ID=5216
For Nobel accounts g
ECONOMICS 302
Second in-class test
Bo E. Honor
Fall 2013
Instructions:
You have 75 minutes to do the following 4 problems.
The total number of points is 75.
The problems have not been ordered in terms of di culty. Note that many of the
questions within ea
Selected Midterm Solutions for Eco 302
October 22, 2013
2009 Midterm
Midterm 2009 Q1.
For the rst question recall the denition of a CI:
[ 1.96 , + 1.96 ]
Plugging in for 3 this just becomes:
[3.589 1.96 2.338, 3.589 + 1.96 2.338] = [.993, 8.17]
For the
ECO302 Final 2008 Solutions
January 20, 2013
Problem 1
(a) t-statistic for 0 and standard error for 4 for Region 1. The t-stat is calculated as
t0 =
0
1.061
= 5.184
=
0.205
se(0 )
So we reject the null that 0 = 0 at a high level of condence (1%). The st
ECONOMICS 312
Final Examination
Bo Honor,
e
Fall 2004
Instructions:
You have 3 hours to do the following 9 problems.
The total number of points is 180.
The problems have not been ordered in terms of diculty. Note that many of the
questions within each
ECONOMICS 302
Second in-class test
Bo Honor
Fall 2010
Instructions:
You have 75 minutes to do the following 4 problems.
The total number of points is 75.
The problems have not been ordered in terms of di culty. Note that many of the
questions within each
ECONOMICS 312
Final
Bo Honor
Fall 2008
Instructions:
You have 180 minutes to do the following 6 problems.
The total number of points is 180.
The problems have not been ordered in terms of di culty.
There is huge variability in the di culty of the question
ECONOMICS 302
First in-class test
Bo Honor
Fall 2010
Instructions:
You have 75 minutes to do the following 4 problems.
The total number of points is 75.
The problems have not been ordered in terms of di culty. Note that many of the questions
within each p
ECONOMICS 312
Final
Bo Honor
Fall 2009
Instructions:
You have 180 minutes to do the following 8 problems.
The total number of points is 180.
The problems have not been ordered in terms of di culty.
There is huge variability in the di culty of the question
ECONOMICS 302
Final
Bo Honor
Fall 2010
Instructions:
You have 180 minutes to do the following 8 problems.
The total number of points is 180.
The problems have not been ordered in terms of di culty.
There is huge variability in the di culty of the question
Fall 2009 - Final Solutions
November 30, 2013
Problem 1.
(a) The formula for a condence interval is
1.96 SE()
Plugging in those numbers here,
[.0669, .0811]
(b) The formula for a t-test is given by,
t=
H 0
SE()
Plugging in,
.0022409 .001
.0003181
10.2
Fall 2010 - Final Solutions
January 16, 2014
Problem 1.
For part (a):
x
1
2
3
4
5
Formula
t
/t
/
1.96
+ 1.96
Written
for Eco 302 at Princeton University, 2014.
straiber@princeton.edu
Number
.0595
.141
7.4
.380
.652
Please send typos, errors, ques
ECO302 Review Session 1 Answer Key
Prepared by: Daniel Goetz
January 14, 2013
1
Regression with one regressor
1.1
a. The standard error of the regression is dened as
SER =
SSR
n2
Since we dont have the SSR, we must compute it. There are two ways we could
Linear Regression
Chp. 69 (minus 7.4) of Stock and Watson
Bo E. Honor
e
Princeton University
2013
Bo Honor (Princeton University)
e
Regression
2013
1 / 46
Multivariate Regression
What if we want to have more than one explanatory variable:
Yi = 0 + X1i 1 +
Some useful distributions
Bo E. Honor
e
Princeton University
Fall, 2013
Bo Honor (Princeton University)
e
Some distributions
Fall, 2013
1/6
Normal Distribution
Fact: For any and any 2 > 0
f x; ,
2
=
1
22
e
(x ) 2
2
is a density.
If X has this density, th
ECONOMICS 302
First in-class test
Bo E. Honor
Fall 2013
Instructions:
You have 75 minutes to do the following 4 problems.
The total number of points is 75.
The problems have not been ordered in terms of di culty. Note that many of the questions within eac
Time Series Topics
Bo E. Honor
e
Princeton University
2013
Bo Honor (Princeton University)
e
Time Series Topics
2013
1/8
Dynamic Causal Eects
Distributed Lags Model
yt = 0 + 1 xt + 2 xt 1 + 3 xt 2 + . . . + r + 1 xt r + u t
Alternative assumptions:
Exogen
Experiments and QuasiExperiments
Bo E. Honor
e
Princeton University
2013
Bo E. Honor (Princeton University)
e
Experiments
2013
1 / 13
Recall the example:
Suppose
yi = 0 + 1 xi + ui
but that ui varies with xi .
Example
yi logwages
xi education
ui ability p
ECO302 Review Session 1
Prepared by: Daniel Goetz
January 13, 2013
Sections
Regression with one regressor
Regression with one regressor (Chapter 4)
Condence intervals and hypothesis testingsingle regressor (Chapter 5)
Multiple regression
Intro to reg
Discrete Choice Models
Bo E. Honor
e
Princeton University
2013
Bo Honor (Princeton University)
e
Discrete Choice
2013
1 / 39
Examples:
To work or not
Vote for Cory Booker or Steve Lonegan
The grade of a midterm
To buy or sell a stock
Which brand to choose
Linear Regression
Chp. 69 (minus 7.4) of Stock and Watson
Bo E. Honor
e
Princeton University
2013
Bo Honor (Princeton University)
e
Regression
2013
1 / 74
Multivariate Regression
What if we want to have more than one explanatory variable:
Yi = 0 + X1i 1 +
Jue Weng
11/16/2013
ECO 302 Assignment 7
Problem 1.
(b)
When (
The 2SLS estimator of ( is (0.9907, 0.9859).
(c)
Case 1: When (
The 2SLS estimator of ( is (0.9161, 0.9698).
Case 2: When (
The 2SLS estimator of ( is (0.8949, 0.6134).
Case 3: When (
The 2SLS
Panel Data
Bo E. Honor
e
Princeton University
2013
Bo Honor (Princeton University)
e
Panel Data
2013
1/4
Example
FatalityRate = 2.01 + 0.15 BeerTax
(0.15)
(1)
(0.13)
FatalityRate = 1.86 + 0.44 BeerTax
(0.11)
(2)
(0.13)
Counterintuitive. Perhaps excluded v
Instrumental Variables Regression
Bo E. Honor
e
Princeton University
2013
Bo E. Honor (Princeton University)
e
IV Estimation
2013
1 / 37
Suppose
yi = 0 + 1 xi + ui
but that ui varies with xi .
Example
yi logwages
xi education
ui ability plus random stu
Bo
Quick Stata Guide
by Liz Foster
Table of Contents
Part 1: Top Ten Stata Commands
describe
generate
regress
scatter
sort
summarize
table
tabulate
test
ttest
1
1
1
3
4
5
5
6
8
10
11
Part 2: Prefixes and Notes
by var:
capture
use of the *
explanation of data