Study guide
Set the focus on stuff we have done after the first midterm.
I plan on giving you one more involved question with somewhat longer derivations than
was possible in the midterms, but the rest of the format is similar to the midterms.
Everythi
cfw_ks Lame M Vmdmg ' @
3." ma LNG cfw_KWFLg CSEWWPw umlg Mat CUB) 1 ow
was Egageg ! loccmc 5 [fix] 4 O - Momma! I
a cogs. . \ .
SO new [3 aoossioz SOLLLEM Lo ME Wing/Wk? " (LEA/ MWMM
dark/LES . _ -
'DfMJTea: A; FwsLAMaWZ-wg uemkwLC; Z is gt Var/w m vwiau
Introduction to Causal Inference
Economics 1630: Econometrics I Fall 2015
Margarita Machelett
Descriptive studies are based on observational data. They provide information about the relationships between
variables but do not answer if one variable change
On Penalization in Regression
Nickolai Riabov
December 8, 2015
Summary
Let the model be given by
Y = X +
Up until this point, weve focused on missing information as the only form of model misspecification.
Under endogeneity, for example, E[x0i ] 6= 0, wh
Economics 1630: Econometrics I
Fall 2015, TTh 1:002:20 pm in Wilson 302
Instructor: Florian Gunsilius
Email: Florian [email protected] (put Econ 1630 in the subject line)
Office Hours: Wednesdays 1.302.30 pm and Thursdays 2.303.30 pm in Robinson
Hall ba
Examples for questions
GLS
Assume we have the following standard linear model
Y = X + ,
where E() = 0, but we have heteroskedasticity. Specifically,
Var() = 2 ,
where is a diagonal matrix (i.e. no serial correlation).
Is OLS still BLUE? Explain.
Answer N
DISCLAIMER: This short note is NOT a relevant subject for this class and nothing
is lost if you never read it. I just wrote this to provide you with another view on the least
squares ordinary estimator in case you are interested. I have never seen it cove
ECON 1630 Fall 2015
PS 1
September 24, 2015
Due Friday, October 2nd at 3.30 pm in the homework drop-off box in Robinson Hall basement.
You may collaborate with classmates, but everyone needs to hand in an own answer. Please
include a print-out of your STA
ECON 1630 Fall 2015
PS 3
November 5, 2015
Due Friday, November 13th at 3.30 pm in the homework drop-off box in Robinson Hall basement. You may collaborate with classmates, but everyone needs to hand in an own answer.
This PS is worth 115 points, but 110 p
ECON 1630 Fall 2015
PS 4
December 1, 2015
Due Friday, December 11th at 3.30 pm in the homework drop-off box in Robinson Hall
basement. You may collaborate with classmates, but everyone needs to hand in an own
answer.
Problem 1[35 points] (Heteroskedastici
(Fancy) Example: Performing a general F -test
Suppose we are given some data (Yi , Xi ), i = 1, . . . , n, and we assume they follow the standard
model:
Yi = 0 + 1 Xi + i
with our standard assumptions on the error. Now suppose we want to test
vs H1 : 0 +
-name: <unnamed>
log: /Users/Yashil/Desktop/ECON 1630/Problem Sets/Problem Set 2/
Econ1630_PS2.log
log type: text
opened on: 12 Oct 2015, 16:40:58
.
end of do-file
. do "/var/folders/th/r2q29r594413f0bvs0h14w580000gn/T/SD11174.000000"
. use "ps1.dta", cle
Meu - - I I @
n.
cfw_MLMW; M A Lg, 0. ML; 34 Sf: w xvx _ , _
Ox Cl, 915 14- H FHSL 430 L414 M4 Jags/cfw_m OMCK
A _-.:.7 Gt 0. (1&3 raga w Rlm. GLOWM )eolb W t
(13. x
W: A '57 A INS Yaw Mum/64
a ,1 _ _. - , Q
w W (A! 4 =9 4% is com luau/irr-
M
h a MM-