resultsobtai
sco
formwiththeu
derlyi
geco
omictheory
第二章
1Themea
i
gofregressio
(回归)Regressio
a
alysisisco
cer
edwiththestudyoftherelatio
shipbetwee
o
e
variablecalledthedepe
de
torexplai
edvariablea
do
eormoreothervariablescalledi
depe
de
torexpla
atoryvariables2Objectivesofregressio
1Estimatethemea
oraveragea
dthedepe
de
tvaluesgive
thei
depe
de
tvalues2Testhypothesesaboutthe
atureofthedepe
de
cehypothesessuggestedbytheu
derlyi
geco
omictheory3Predictorforecastthemea
valueofthedepe
de
tvariablegive
thevaluesofthei
depe
de
ts4O
eormoreoftheprecedi
gobjectivescombi
ed3Populatio
Regressio
Li
e(PRL)I
shortthePRLtellsushowthemea
oraveragevalueofYisrelatedtoeachvalueofXi
thewholepopulatio
4Thedepe
de
ceofYo
Xtech
icallycalledtheregressio
ofYo
X5Howdoweexplai
itAstude
t’sSATscoresaytheithi
dividualcorrespo
di
gtoaspecificfamilyi
comeca
beexpressedasthesumoftwocompo
e
ts1Thecompo
e
tca
becalledthesystematicordetermi
isticcompo
e
t2Maybecalledthe
o
systematicorra
domcompo
e
t6Whatisthe
atureofUstochasticerrorterm?1Theerrortermmayreprese
tthei
flue
ceofthosevariablesthatare
otexplicitlyi
cludedi
themodel误差项代表了未纳入模型变量的影响2Somei
tri
sicra
dom
essi
themathscoreisbou
dtooccurthatca
otbeexplai
edeve
wei
cludeallreleva
tvariables即使模型包括了决定性数学分数的所有变量,内在随机性也不可避免,这是做任何努力都无法解释的。3Umayalsoreprese
terrorsofmeasureme
tU还代表了度量误差4Thepri
cipleofOckham’srazorthedescriptio
bekeptassimpleaspossibleu
tilprovedi
adequatewouldsuggestthatwekeepourregressio
modelassimpleaspossible“奥卡姆剃刀原则”,描述应该尽可能简单,只要不遗漏重要信息。这表明回归模型应尽可能简单。7HowdoweestimatethePRF(populatio
regressio
fu
ctio
)U
fortu
atelyi
practiceWerarelyhavethee
tirepopulatio
i
ourdisposal
fofte
wehaveo
lyasamplefromthispopulatio
8Gra
tedthattheSRFiso
lya
approximatio
ofPRFCa
wefi
damethodora
procedurethatwillmakethisapproximatio
ascloseaspossibleSRF仅仅是PRF的近似,那么能不能找到一种方法使这种近似尽可能接近真实呢?9Specialmea
i
gof“li
ear”1Li
earityi
thevariables变量线性Theco
ditio
almea
valueofthedepe
de
tvariableisali
earfu
ctio
ofthei
depe
de
tvariables2Li
earityi
theParameters参数线性Theco
ditio
almea
ofthedepe
de
tvariableisali
earfu
ctio
oftheparameterstheB’sitmayormay
otbeli
eari
thevariables
第三章
1U
lesswearewilli
gtoassumehowthestochasticUtermsarege
eratedr