The first part of this assignment will be due Monday, March 13 and will be submitted in class and online. The second section of the assignment will be submitted online by Friday, March 17. Make sure...

SAGINAW VALLEY STATE UNIVERSITY, DEPARTMENT OF ECONOMICS ECON 335: APPLIED STATISTICS, WINTER 2017 DR. MICAH DELVECCHIO [email protected] 311 CURTISS HALL Take-Home Assignment #2 Section I due in-class Monday, March 13, 2017 Section II due in Canvas (online), Friday, March 17, 2017 INSTRUCTIONS The ?rst part of this assignment will be due Monday, March 13 and will be submitted in class and online. The second section of the assignment will be submitted online by Friday, March 17. Make sure to write-up (or type-up) the ?rst section. In the second section, you will need to use the Excel ?les posted in Canvas and submit two ?les in that same assignment link. One ?le will be the .xlsx extension that shows your work, and another with a .csv extension which only contains answers in the appropriate cells. The .csv ?le is the most important. Your printed versions should have your name, and section number on them. SECTION I: CONCEPTUAL/WRITTEN (25 POINTS) CHAPTER 13: SIMPLE LINEAR REGRESSION #1 In what ways does a linear regression give you more information than comparing means or an ANOVA? #2 Write a description as indicated in each part below. The type of description varies but all will show a violation of one of the 4 regression assumptions discussed in sec- tions 13.4 and 13.5. a) Write an equation which violates the ?rst assumption: linearity. b) Illustrate a graph which shows a violation the second assumption: indepen- dence of the error terms. This is also known as autocorrelation of the residuals, in case you want to ?nd it online. c) If the residuals are not normally distributed, what might the distribution look like? That is, in which parts of the distribution of the residuals would you most likely ?nd the values clustered? d) Illustrate a graph which violates the fourth assumption: equal variance of the error terms. This is also known as heteroscedasticity, in case you want to ?nd it online. #3 Refer to the following population regression model to answer this...