# Register now to access 7 million high quality study materials (What's Course Hero?) Course Hero is the premier provider of high quality online educational resources. With millions of study documents, online tutors, digital flashcards and free courseware, Course Hero is helping students learn more efficiently and effectively. Whether you're interested in exploring new subjects or mastering key topics for your next exam, Course Hero has the tools you need to achieve your goals.

5 Pages

### 2008sol1

Course: ECON 620, Spring 2008
School: Cornell
Rating:

Word Count: 774

#### Document Preview

University Cornell Department of Economics Econ 620 Spring 2007 Instructor: Professor Kiefer TA: Jae-ho Yun (jy238@cornell.edu) Problem Set # 1: Solution Key 1. b 1 X (xi x)( + xi + x2 + &quot;i ) x)yi 0 1 2 i = (xi x)2 (xi x)2 X (xi x)x2 X (xi x) X (xi x)xi i + 2 = + 1 0 (xi x)2 (xi x)2 (xi x)2 X (xi x)&quot; + (xi x)2 X (xi x)x2 X (xi x)&quot; i = + 1+ 2 (xi x)2 (xi x)2 = Note: X (xi (xi X (xi x) =...

Register Now

#### Unformatted Document Excerpt

Coursehero >> New York >> Cornell >> ECON 620

Course Hero has millions of student submitted documents similar to the one
below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.

Course Hero has millions of student submitted documents similar to the one below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.
University Cornell Department of Economics Econ 620 Spring 2007 Instructor: Professor Kiefer TA: Jae-ho Yun (jy238@cornell.edu) Problem Set # 1: Solution Key 1. b 1 X (xi x)( + xi + x2 + "i ) x)yi 0 1 2 i = (xi x)2 (xi x)2 X (xi x)x2 X (xi x) X (xi x)xi i + 2 = + 1 0 (xi x)2 (xi x)2 (xi x)2 X (xi x)" + (xi x)2 X (xi x)x2 X (xi x)" i = + 1+ 2 (xi x)2 (xi x)2 = Note: X (xi (xi X (xi x) = 0; x)2 (xi X (xi x)xi =1 x)2 X (xi Take the conditional expectation; E( b 1 jxi ) = + x)x2 i x)2 1 2 (xi Therefore, b 1 is biased. V ar( b 1 jxi ) X (xi Note: E( (xi x)" j xi ) = 0 x)2 Next, the conditional variance is; = E[( b 1 E( b 1 jxi ))2 j xi ] X (xi x)" = E[( )2 j xi ] (xi x)2 (cross product terms are all zero) 2 = P (xi 1 x)2 2. i) The correlation between y and z is 1 if a>0, -1 if a<0 and 0 if a=0. ii) The support of the joint distribution of y and z can be represented by the following set; f(x; y) : z = ay; y 2 Rg (It is a straight line) iii) The correlation between y and x ( x = y 2 ) is zero, since Cov(y; y 2 ) = E(y y 2 ) E(y)E(y 2 ) = E(y 3 ) = 0 iv) The support is f(x; y) : x = y 2 ; y 2 Rg v) In the above problem, clearly, there are dependence between x and y. However, our linear relationship measure, correlation does not capture this relationship. 3. (a) OLS is BLUE if "i are homoskedastic with mean zero. 1 1 ( 1) + ( 1) = 0 for i=1,2 2 2 1 1 ( 1)2 + ( 1)2 = 1 for i=1,2 E("2 ) = i 2 2 : Cov("i ; "j ) = E("i "j ) E("i )E("j ) = 0 for i 6= j E("i ) = Mean: Variance: Covarince Hence, our OLS is BLUE. (b) b P P xy xi (xi + "i ) P i2i = P 2 = = xi xi 1 "1 + 2 "2 = 1+ 12 + 22 P xi "i + P 2 xi (c) The alternative estimator We have the following exact distribution of b Prob. :25 :25 :25 :25 b 2=5 4=5 6=5 8=5 P P y (xi + "i ) P i = P = xi xi "1 + "2 = 1+ 1+2 = P "i +P xi is clearly unbiased. Hence, the exact distribution of is Prob. :25 :50 1=3 1 :25 5=3 2 (d) The exact variance of the two estimators: "1 + 2"2 1 1 )= (V ("1 ) + 4V ("2 )) = 5 25 5 "1 + "2 1 2 = V( ) = (V ("1 ) + V ("2 )) = 3 9 9 V( Hence, V ( V (b) = V( ) 4. (a) Recall that ) > V ( ); b which is consistent with b being BLUE. b X (xi x)(yi y) (xi x)2 = and X (xi Hence, x)(yi X (xi = = = y bx X y) = xi yi nxy X x)2 = x2 nx2 i b (b) R2 is de...ned as the ratio of the explained sum of squares(ESS) to total sum of squares(TSS). P 2 b 2 P(xi x)2 2 xi P =b P 2 (yi y)2 yi b (n b b 30 = 0:5 60 440 0:5 22 220 = 15 22 R2 = nx2 60 = 0:52 8900 22 ny 2 202 = 0:15 (c) By the normality assumption, we know that 2 Moreover, N( ; P (xi 2)S 2 2 2 x)2 ) (n 2) P 2 where S2 = n 1 2 ei : We can also show that b and S 2 are independent each other. Then, 3 r r P b 2 x)2 2)S 2 2 (xi = (n (n 2) pP b S (xi x)2 t(n 2) We want to reject the null hypothesis if b t= under the null hypothesis. On the other hand, 1 n 1 n 2 4:25 2 X X 1 n 2 X 2 pP S (xi x)2 > t0:975 (20) S2 = = = e2 = i (yi 2 (yi + b 2 + b x2 i 2b yi + 2b b xi b b xi )2 2 b xi yi )2 Hence, the test statistics is given by 0:5 t= q 0 4:25 60 = 1:8787 Since t0:975 (20) = 2:086;we cannot reject the null hypothesis. 5. i) b1 0 0 (X1 M2 X1 ) 1 X1 M2 y 0 0 (X1 M2 X1 ) 1 X1 M2 (X1 1 + X2 2 + ") 0 1 0 0 0 (X1 M2 X1 ) X1 M2 X1 1 + (X1 M2 X1 ) 1 X1 M2 X2 0 0 +(X1 M2 X1 ) 1 X1 M2 " 0 1 0 = X 1 M2 " 1 + 0 + (X1 M2 X1 ) = = = 2 Take expectation. E(b1 ) = = = = 1 0 0 + E[(X1 M2 X1 ) 1 X1 M2 "] 0 1 0 X1 M2 E[&q...

Find millions of documents on Course Hero - Study Guides, Lecture Notes, Reference Materials, Practice Exams and more. Course Hero has millions of course specific materials providing students with the best way to expand their education.

Below is a small sample set of documents:

Cornell - ECON - 620
Cornell University Department of Economics Econ 620 Spring 2008 TA: Jae Ho YunSolution Key for PS #41. Suppose that the regression model is;y = X +&quot; E(&quot;) = 0; E(&quot;0 ) = Assume that is known.2(a) the covariance matrix of the OLS estimator bOL
Cornell - ECON - 620
1Economics 620, Lecture 8a: Asymptotics II Uses of Asymptotic Distributions: Suppose Xn ! 0 in probability. (What can be said about the distribution of Xn ?) In order to get distribution theory, we need to norm the random variable; we usually look
Cornell - ECON - 620
Economics 620, Lecture 9: Asymptotics III: Maximum Likelihood EstimationNicholas M. KieferCornell UniversityProfessor N. M. Kiefer (Cornell University)Lecture 9: Asymptotics III(MLE)1 / 20Jensen Inequality : Suppose X is a random variable w
Cornell - ECON - 620
Economics 620, Lecture 12: Generalized Least Squares (GLS) IINicholas M. KieferCornell UniversityProfessor N. M. Kiefer (Cornell University)Lecture 12: GLS II1 / 19Example 1: Grouping (or data on averages)Suppose the &quot;true&quot; model is y = X
Cornell - ECON - 620
Economics 620, Lecture 13: Time Series INicholas M. KieferCornell UniversityProfessor N. M. Kiefer (Cornell University)Lecture 13: Time Series I1 / 19AUTOCORRELATIONConsider y = X + u where y is T T 1. 1, X is T K, is K 1 and u isWe are
Cornell - ECON - 620
Economics 620, Lecture 14: Lags and DynamicsNicholas M. KieferCornell UniversityProfessor N. M. Kiefer (Cornell University)Lecture 14: Lags and Dynamics1 / 15Modern time-series analysis does not treat autocorrelation as necessarily a proper
Cornell - ECON - 620
Economics 620, Lecture 17: SEM MiscellaneousNicholas M. KieferCornell UniversityProfessor N. M. Kiefer (Cornell University)Lecture 17: SEM III1 / 15Last time, we saw that the 2SLS estimator is given by ^ = (Z 0 MZ ) 1 Z 0 My1 . Note that th
Cornell - ECON - 620
Economics 620, Lecture 18: Nonlinear ModelsNicholas M. KieferCornell UniversityProfessor N. M. Kiefer (Cornell University)Lecture 18: Nonlinear Models1 / 18The basic point is that smooth nonlinear models look like linear models locally. Mod
Cornell - ECON - 620
1Economics 620, Lecture 8: Asymptotics IWe are interested in the properties of estimators asn ! 1:Consider a sequence of random variablesfXn; n1g:Often Xn is an estimator such as a sample mean or c n Often it is convenient to center the s
Cornell - ECON - 620
1Lecture 11: Generalized Least Squares (GLS)In this lecture, we will consider the model y = X + &quot; retaining the assumption Ey = X .However, V (&quot;) = V where V = 2we no longer have the assumption V (y ) = 2 I . Instead we add the assumption V (
Cornell - ECON - 620
1Lecture 12: GLS IIExample 1: Grouping (or data on averages)Suppose the true model is y = X + &quot; where E&quot; = 0 and V (&quot;) = 2I . Suppose that the available data are [~ X ] arranged in y ~ M (&gt; K ) groups and the group means are the only givens (e.
Cornell - ECON - 620
1LECTURE 13: TIME SERIES I AUTOCORRELATION : Consider y = X + u where y is T is K 1 and u is T 1. 1, X is TK,We are using T and not N for sample size to emphasize that this is a time series. The natural order of observations in a time series sug
Cornell - ECON - 620
1Lecture 17: SEM MiscellaneousLast time, we saw that the 2SLS estimator is given by ^ = ( Z 0 M Z ) 1 Z 0 M y1 . Note that the 2SLS method is a limited information method.In this case, what are the sources of ine ciency?1. 2SLS does not explo
Cornell - ECON - 620
1Lecture 18: Nonlinear ModelsThe basic point is that smooth nonlinear models look like linear models locally. Models linear in parameters are no problem even if they are nonlinear in variables. For example'(y ) = 0 + 1 1(x1) + 2 2(x2) + :with
Cornell - ECON - 620
Cornell University Department of Economics Econ 620 Spring 2008 Instructor: Professor Kiefer TA: Jae Ho YunProblem Set #4(Due: April 8, Tuesday) 1. Suppose that the regression model is;y = X +&quot; E(&quot;) = 0; E(&quot;0 ) = Assume that is known.2(a) Wh
Cal Poly - CS - 4744
654xmax / h321000.20.40.60.811.22
Cal Poly - CS - 4744
0.70.60.5 x 2 / 2 h0.40.30.20.10 0.911.1 (Angle in radians)1.21.31.41.51.6
Cal Poly - CS - 4744
20log (Function Value)-2-4-6-8-10-12050100150200250300350Iteration
Cal Poly - CS - 4744
Cal Poly - CS - 4744
Cal Poly - CS - 4744
Cal Poly - CS - 4744
1030 1020 Size of the Universe (meters) 1010 100 10-10 10-20 10-30 10-40Age of the Universe (years) 10-50 10-40 10-30 10-20 10-10 100101010-40 10-30 10-20 10-10 100 1010 Age of the Universe (seconds)1020
Age of the Universe (seconds) 10-40 10-30 10-20 10-10 100 1010 Mass-Energy Density of the Universe (kg/m3) 10100 1090 1080 1070 1060 1050 1040 1030 1020 1010 100 10-10 10-20 10-30 10-50 10-40 10-30 10-20 10-10 100 Age of the Universe (years)10201
UMass Lowell - CS - 570
Scribe: Swati Gupta Date: 09/09/02 Lecture Notes - 1Authentication ApplicationsTwo main topics: 1) Kerberos, 2) X.509 From the previous course (91.457), we have learned: 1. Cryptography Symmetric (secret) key cryptographic algorithms Examples of
UMass Lowell - CS - 570
Scribe: Yan Liu Date: 09/16/02 Lecture Notes - 2X.509 &amp;PGPTwo main topics: 1) Kerberos 2) X.509X.5091. X.509 use public key encryption (such as RSA) and DSS for authentication. For example: Alice created a pair of keys: Public key(KUA) &amp; Privat
UMass Lowell - CS - 570
Scribe: Patrick Hoey Date: 11-November-2002 Lecture Notes 9 91.570 Computer Security II Prof. Wang Fall 2002 Notes about setup in labs for projects: Linux machines are setup in room OS 305. Follow chapter 6 to setup FreeS/WAN. 3 machines are running
UMass Lowell - CS - 570
Scribe: Fang Wu Date: 11/18/02Topics: I. PPTP Protocol II. Building a PPTP VPN ClientI. PPTP Protocol General Info.PPTP, the Point-to-Point Tunneling Protocol, is an enhancement to PPP developed by Microsoft. PPTP is fairly well integrated int
UMass Lowell - CS - 570
91.570 Computer Security II Scribe: Jie Lu Date: 12/09/02Advanced CISCO Router SecurityACL traditional, standard feature Stateless packet filtering Advanced features Dynamic ACL (DACL) Time-based ACL (TACL) Reflexive ACL (RACL) Context Based
Iowa State - IE - 361
Distributions Exam 2Distributions E1+E220304050607080906080100120140160180Quantiles100.0% 99.5% 97.5% 90.0% 75.0% 50.0% 25.0% 10.0% 2.5% 0.5% 0.0% maximum 84.000 84.000 84.000 80.500 67.750 51.500 37.250 31.500 28.0
Iowa State - IE - 361
IE 361 SyllabusInstructor: Steve Vardeman 304-D Snedecor/2078 Black 294-2535 vardeman@iastate.edu http:/www.public.iastate.edu/~vardeman/ Official Office Hours: 2-3 MWF (304-D Snedecor) Grader: To be announcedRequired Text: Statistical Quality Ass
Iowa State - IE - 361
IE 361 Homework Set #1 Fall 2000 1. Do problem 2.5 of the text using the range-based formulas. Then redo it using the ANOVA-based formulas on page 27 of the text. 2. Below is a data set from a real calibration study (taken from a paper by John Mandel
Iowa State - IE - 361
IE 361 Homework Set #1 Spring 2004 1. A group doing the in-class measurement study on foam peanuts obtained the following results: i) Single measurements on 8 ) different peanuts had sample mean C &amp;* inch and sample standard deviation =C !# inch.
Iowa State - IE - 361
IE 361 Homework Set #1 Fall 20041. An IE 361 in-class measurement demonstration (using a fairly nice micrometer) produced size measurements for inexpensive binder clips. (The raw data are on the IE 361 Handout Page for anyone who is interested.) The
Iowa State - IE - 361
IE 361 Homework Set #1 Fall 2002 1. One of the groups doing the in-class measurement study on foam peanuts obtained the following results: i) Single measurements on 8 oe ) different peanuts had sample mean C oe &amp;* inch and sample standard deviation =
Iowa State - IE - 361
IE 361 Class Outline, Fall 2001 Introduction/Chapter 1 (&amp;Chapter 9) quality defined oe _ quality of design quality of conformance statistics defined oe _ processes vs. products customer focus continual improvement 6-step process-oriented quality assu
Iowa State - IE - 361
IE 361 Class Outline, Fall 2002 Introduction/Chapter 1 (&amp;Chapter 9) quality defined oe _ quality of design quality of conformance statistics defined oe _ processes vs. products customer focus continual improvement 6-step process-oriented quality assu
Iowa State - IE - 361
IE 361 Class Outline, Fall 2000 Introduction/Chapter 1 (&amp;Chapter 9) quality defined _ quality of design quality of conformance statistics defined _ processes vs. products customer focus continual improvement 6-step process-oriented quality assuranc
Iowa State - IE - 361
IE 361 SyllabusInstructor: Steve Vardeman 304-D Snedecor/2078 Black 294-2535 vardeman@iastate.edu http:/www.public.iastate.edu/~vardeman/ Official Office Hours: 2-3 MWF (304-D Snedecor) Teaching Assistant: Dewi Rahardja 3011 Black gaby@iastate.edu
Iowa State - IE - 361
IE 361 Exam 1 October 5, 2004 Prof. Vardeman1. IE 361 students Demerath, Gottschalk, Rodgers and Watson worked with a manufacturer on improving the consistency of several critical dimensions of a part. One of their first steps toward process improv
Iowa State - IE - 361
IE 361 Exam 2 November 10, 2004 Prof. Vardeman 1. Limestone blocks of length approximately l = 1 m , width approximately w = .5 m , and height approximately h = .25 m are used in a construction project. The actual sizes of the blocks and the density,
Iowa State - IE - 361
IE 361 Exam 3 December 15, 2004 Prof. VardemanAn IE 361 project group worked with a plastics manufacturer and experimented on an extrusion process. They conducted what was ultimately a (partially replicated) 26 - 2 fractional factorial study aimed
Iowa State - IE - 361
IE 361 Schedule Fall 2006Date T 8/22 Web Viewing (Preferably Before Class) M1 Class Activity Lab1 (1st Measurement Study) Lab2 (2nd Measurement Study) Lab 3 (Gauge R&amp;R Calculations) Lab 4 (Calibration Study) Lab 5 (R&amp;R in a Go/No-Go Study) Lab 6 (De
Iowa State - IE - 361
IE 361 SyllabusInstructor: Steve Vardeman 304-D Snedecor/2078 Black 294-2535 vardeman@iastate.edu http:/www.public.iastate.edu/~vardeman/ http:/www.public.iastate.edu/~vardeman/IE361/ie361vard.html Official Office Hours: MWF 2:30-3:30 (304-D Snedeco
Iowa State - IE - 361
IE 361 SyllabusInstructor: Steve Vardeman 304-D Snedecor/2078 Black 294-2535 vardeman@iastate.edu http:/www.public.iastate.edu/~vardeman/ http:/www.public.iastate.edu/~vardeman/IE361/ie361vard.html Official Office Hours: MWF 2:30-3:30 (304-D Snedeco
Iowa State - IE - 361
Iowa State - IE - 361
IE 361 In-Class Gage R&amp;R Demonstration Fall 2000MTB &gt; Print 'Part' 'Operator' 'y'. Data Display Row 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Part 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 Operator 1 1 2 2 3 3 4 4 1 1 2 2
Iowa State - IE - 361
IE 361 Gage R&amp;R Study Fall 2001MTB &gt; Print c1-c3 Data Display Row 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Part 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 Operator 1 1 2 2 3 3 4 4 1 1 2 2 3 3 4 4 1 1 2 2 3 3 4 4 Measuremen
Iowa State - IE - 361
IE 361 Homework 1 1. a. t distribution: y +/- t (s/m.5) .537 (+/-) 1.476(.008/ 6.5) = [.532, .542] 2 distribution: [(m-1)s2/2 m-1, upper ).5 , (m-1)s2/2 m-1, lower ).5 ] [(5(.0082)/11.07).5 , (5(.0082)/1.145).5] = [.00538, .0167] These two quantitie
Iowa State - IE - 361
IE 361 SyllabusInstructor: Steve Vardeman 304-D Snedecor/2078 Black 294-2535 vardeman@iastate.edu http:/www.public.iastate.edu/~vardeman/ Official Office Hours: TWRF 2 (304-D Snedecor) Grader: To be announcedRequired Text: Statistical Quality Assu
Iowa State - IE - 361
Iowa State - IE - 361
Statistics and Measurement (Understanding and Quantifying Measurement Uncertainty)The general question to be addressed here is &quot;How do statistical methods inform measurement/metrology?&quot; Some answers will be phrased in terms of methods for separating
Iowa State - IE - 361
IE 361 Lab#1/ 1st Measurement Study_ (Name) Before making each measurement, close the jaws of the caliper and reset the &quot;zero&quot; on the dial. - First, measure &quot;the size&quot; of each foam nugget in your bag once. Let everyone at your table do the same. Th
Iowa State - IE - 361
IE 361 Lab#10/ Two-Way Analysis of a Paper Helicopter Experiment _ (Name) Below are some data collected by one team in the helicopter study of Lab #9. This lab consists of making a two-way analysis of these data. Recall that in Lab #9 we called the h
Iowa State - IE - 361
IE 361 Lab#11/ Analysis of 2 p Factorial and 2 p - q Fractional Factorial Data _ (Name) An example of a partially replicated 23 factorial experiment due originally to R. Snee and appearing in Engineering Statistics by Hogg and Ledolter concerns the e
Iowa State - IE - 361
IE 361 Lab#2/ 2nd Measurement Study_ (Name)Before handing in Lab #1, record below from the top of page 2 on Lab #1 the summary statistics that you report. From the &quot;first sample&quot; of n = 8 measurements on different peanuts: y = __s y = _From t
Iowa State - IE - 361
IE 361 Lab#5/ R&amp;R for &quot;0/1&quot; or Go/No-Go Data_ (Name)We are going to collect and make simple analyses of some &quot;0/1&quot; or &quot;Go/No-Go&quot; data. Organize yourselves into groups of 4 or 5 for this exercise. One of your group members needs to serve as the re
Iowa State - IE - 361
IE 361 Lab#7/ Attributes Control Charts_ (Name)In this lab, we're going to do some collection of attributes data and corresponding control charting. Our &quot;production process&quot; is going to be a &quot;coin spinning process.&quot; That is, we're going to spin c
Iowa State - IE - 361
IE 361 Lab#8/ 1st Paper Helicopter Study and Process Capability _ (Name) In this lab, we're going to collect some &quot;paper helicopter flight time data&quot; for two purposes. First, we'll do this as practice with the mechanics of making, flying, and timing
Iowa State - IE - 361
IE 361 Lab#9/ A One-Way Analysis of a Paper Helicopter Experiment _ (Name) In this lab, we're going to do some experimenting with different &quot;paper helicopter designs.&quot; The mechanics of our testing will be exactly as in Lab #8. That is, when your heli
Iowa State - IE - 361
Page 1FRACTIONAL FACTORIAL (#:; ) STUDIESMotivation: For : factors, even #: gets big fast for : &quot;! #: &quot;!#% Example A B C D E F G H J K L M N O P Hendrix 1979 Chemtech Coating Roll Temp Solvent Polymer X-12 Preheat Web Type Coating Roll Tension N