Economics 240B
Winter 2012
Professor Startz
Final Exam
Name:
Student number:
.
This exam is designed to take 2 hoursso you can have 2.5 hours. (No, not three.)
Answer any 7 out of the 8 questions.
Each question is worth 20 points, although some questions
Section B05 Nicholas Nguyen 5.1 a) > v = [2;0;-1];w=[1;3;3];x=[6;1;-3];y=[1;0;2];z=[2;-15;-1]; > y'*x ans = 0 > y'*z ans = 0 > x'*z ans = 0 So the set cfw_x, y, z are all orthogonal to each other. The maximum number of non-zero orthogonal vectors that I c
Economics 240B
Winter 2013
Professor Startz
Final Exam
This exam is designed to take 2 hoursso you can have 2.5 hours. (No, not three.)
Please answer all six questions. Each question is worth 25 points, although some questions are easier
than others. Be s
Chapter Seven
Revealed Preference
Revealed Preference Analysis
Suppose we observe the demands
(consumption choices) that a consumer
makes for different budgets. This
reveals information about the
consumers preferences. We can use
this information to .
Re
Chapter Seven
Revealed Preference
Revealed Preference Analysis
Suppose we observe the demands
(consumption choices) that a consumer
makes for different budgets. This
reveals information about the
consumers preferences. We can use
this information to .
Re
Multiple Regression
Copyright Dick Startz
1
Goals
right hand side variables
= 2 as an example
t-tests involving multiple coefficients
F-tests
Copyright Dick Startz
2
Regression coefficient with two RH S
variables
2 1 1 2 2
2
1 =
2 2
1 2 2
1
2
2
var 1
Limited Dependent Variables
Copyright Dick Startz
1
Goals
Nonlinear estimation
Maximum likelihood
Likelihood function, information matrix, likelihood
ratio test
Limited dependent variables
Probit/logit
Copyright Dick Startz
2
MLE for sample mean
=
Lecture 6 Correlation
Stock example: stock prices are likely to be correlated. Need
a measure of strength of correlation.
Microarray example
Defining correlation :
Procedure of computing correlation
(1)standardize x, (2)standardize y, (3)average
product o
ARIMA
Copyright Dick Startz
1
Goals
AutoRegressive Integrated Moving Average
AutoRegressive models
Lag operator
Moving Average models
ARMA
Autoregressive Integrated Moving Average
Nonstationary models
Copyright Dick Startz
2
Autocorrelation Function
= ,
Economics 240B
Winter 2014
Professor Startz
Homework 4
1. Using the CPS data set, regress log(wsal_val) on a constant and education. What's the
percentage return to another year of education. Now regress wsal_val on a constant and
education, being sure th
ARIMA
Copyright Dick Startz
1
Goals
AutoRegressive Integrated Moving Average
AutoRegressive models
Lag operator
Moving Average models
ARMA
Autoregressive Integrated Moving Average
Nonstationary models
Copyright Dick Startz
2
Turning AR(1) into history of
Schools
Subjects
Standardized Tests
Books
skkobe
My Account
Logout
Explore
Schools
Subjects
Standardized Tests
Books
Study Documents
Find Documents
My Documents
Textbook Help
Upload Documents
Tutors
Ask a Question
My Questions
About Tutors
Fla
Economics 240B
Winter 2014
Professor Startz
Homework 4
1. Using the CPS data set, regress log(wsal_val) on a constant and education. What's the
percentage return to another year of education. Now regress wsal_val on a constant and
education, being sure th
Serial Correlation
Copyright Dick Startz
1
Goals
Error terms across observations are correlated
with one another
Specific model, = 1 +
OLS variance formula wrong
GLS (Cochrane-Orcutt)
Testing for serial correlation
Serial correlation with a lagged
Misspecification
Copyright Dick Startz
1
Goals
Omitted variable.
Correlated with included variables or not.
Theils misspecification theorem.
Included extraneous variable.
Correlated with error or not.
Running a regression backwards.
Copyright Dick S
Limited Dependent Variables
Copyright Dick Startz
1
Goals
Nonlinear estimation
Maximum likelihood
Likelihood function, information matrix, likelihood
ratio test
Limited dependent variables
Probit/logit
Copyright Dick Startz
2
MLE for sample mean
= +
Endogeneity
Copyright Dick Startz
1
Goals
Consequence of RHS variables being
correlated with error term
Instrumental variables (IV)
Errors-in-variables
Copyright Dick Startz
2
Limit of sample variance?
~ 0, 2
2
var =
?2
var =
2 ? 2
lim
=
Copyright
Econometrics
Economics 240B
Professor Dick Startz
Copyright Dick Startz
1
Goals
Illustrate estimation
Forecasting example
Idea of regression coefficients as random
variables
Quick EViews example
Foreshadow technical material
Programming in Matlab
U
Economics 240B
Winter 2013
Professor Startz
Sketched Out Final Answers
Note that not all the details are shown here. The sketches here are just to point you in the right
direction. A good exam answer would have more details.
1)
a)
(
),
(
)
(
)
b)
(
),
(
)
Dif-in-Dif and Fixed Effects
Copyright Dick Startz
1
Goals
Dif-in-Dif as control for unobservables with
repeated cross-sections.
Fixed effects as control for unobservables in
panel data.
Copyright Dick Startz
2
LaborSupply
Chapter9
1
Question
n
n
n
n
A doctor earns $150 an hour
He wants to take 2 hours off from work to see a movie
The ticket costs $8
How many dollars worth of other consumption must he
give up to see the movie?
(A)
(B)
(C)
(D)
0
Cant tell
$8
$308
Chapter Thirty-Five
Information Technology
Information Technologies
The crucial ideas are:
Complementarity
Network externality
Information Technologies;
Complementarity
Definition: Commodity A complements
commodity B if more of commodity A
increases t