计量经济学 习题exercise 1.doc
上传人:sy****28 上传时间:2024-09-15 格式:DOC 页数:3 大小:75KB 金币:15 举报 版权申诉
预览加载中,请您耐心等待几秒...

计量经济学 习题exercise 1.doc

计量经济学习题exercise1.doc

预览

在线预览结束,喜欢就下载吧,查找使用更方便

15 金币

下载此文档

如果您无法下载资料,请参考说明:

1、部分资料下载需要金币,请确保您的账户上有足够的金币

2、已购买过的文档,再次下载不重复扣费

3、资料包下载后请先用软件解压,在使用对应软件打开

AnindyaBanerjeeFirstSemester2009ECONG26/27/28/299October2009ExerciseSheet11.i)Definethesimplelinearcorrelationcoefficient,r,betweentheobservedvaluesand(fori=1,2,....,n)oftwovariablesXandY.ii)SupposeanordinaryleastsquareslinearregressionisestimatedbetweenthevariablesYandX:.Showthatthecoefficientofdeterminationofthisregression,R2,isequaltor2definedini).Henceprovethat-1≤r≤1.iii)NowsupposeanordinaryleastsquareslinearregressionisundertakenbetweenXandY:Whatistherelationshipbetweenestimatedinthisregressionwithestimatedin(ii)?Canwededucetheestimateoffromtheregressionstatisticsin(ii)?iv)Ingeneral,whyisShowunderwhatcircumstanceswillIllustrateyouranswer.2.Thefollowingaredataon:Y=quitrateper100employeesinmanufacturing;X=unemploymentrateYearYXYearYX19901.36.219972.33.619911.27.919982.53.319921.45.919992.73.319931.45.720002.15.619941.55.020011.96.919951.94.020022.25.619962.63.2i)CalculatearegressionofYonX.ii)Constructa95%confidenceintervalforα.iii)TestthenullhypothesisH0:β=0atthe5%significancelevel.iv)Constructa90%confidenceintervalforσ2,whereσ2isthevarianceoftheregressionerror.v)CalculatetheR2oftheregression.vi)Arethereanyreasonswhytheassumptionsoftheclassicalnormallinearmodelareinvalid?3.AtBirminghamUniversity7economicsundergraduateswererandomlyselectedfromthepopulationandsurveyed.Theywereasked:i)Whatwastheaveragenumberofhoursoflecturesperweekyouattendedlastyear?ii)Whatwastheaveragenumberofhoursperweekyouspentdrinkinglastyear?Thefollowingdatawerecollected:StudentHoursoflecturesHoursofdrinking13.6322.21533.1843.5952.71262.61273.94i)Estimatetheordinaryleastsquaresequation,whereYistheaveragenumberofhoursoflecturesandXrepresentstheaveragenumberofhoursdrinkingperweek.ii)CalculatetheR2oftheregressionandthestandarderroroftheequation.iii)TestthenullhypothesisH0:β=0.4.LetYbeatimedependentvariableandconsiderthelinearregressionmodel:wherei)Le