1. 用图表检验Analyze -> regression -> Linear-> Plots
9 o( R1 J+ [* L/ VScatter plot of the standardised residuals on the standardised predicted values (ZRESID as the Y variable, and ZPRED as the X variable( y/ h+ g! Q7 B- s
2 D9 `) x! |$ e2 c. Y8 o$ w% M
如果图表显示有可能存在异方差,需要用统计检验来进一步检测异方差是否确实存在。 N3 _' n6 @4 n. e0 |6 z2 q" W
2. 用统计检验7 f7 i) ?7 d# t/ \
Heteroscedasticity——Testing and Correcting in SPSS.pdf
Gwilym Pryce March 2002.doc
(172.5 KB, 下载次数: 4)
. w/ l# P2 l' _, o, v
r$ p5 K" S Y, v+ r4 `9 U4 S* {8 Q2 }5 |; q
Levene’s Test3 d9 b9 v9 r0 e! W
Goldfeld-Quandt Test8 Y# F3 Y9 m* ]' d& Q+ l7 y
Breusch-Pagan Test$ z4 Q9 ]9 W: h' w% O" k) _
White‘s Test (比较常用来检验异方差)
: E8 z3 s' W5 R& u% cAssume you want to run a regression of wage on age, work experience,education, gender, and a dummy for sectorofemployment (whether employed in the public sector). wage = function(age, workexperience, education, gender, sector) or, as your textbook will have it, wage = b1 + b2*age + b3*work experience+ b4*education + b5*gender + b6*sector The White’s test is usually used as a test for heteroskedasticity. In this test, a regression of the squares ofthe residuals is run on the variables suspected of causing theheteroskedasticity, their squares, and cross products. (residuals)2 = b0 + b1 educ + b2 work_ex+ b3 (educ)2 + b4 (work_ex)2 + b5(educ*work_ex) Y! i2 u7 `+ x
& ^6 ]4 b* g/ {5 t7 p% f. }" q$ I* X1 q
! ^7 i) F0 R) f8 K3 U! a2 p# n5 }% G, }/ A
White’s Test · Calculate n*R2 à R2 = 0.037, n=2016 à Thus, n*R2 = .037*2016 = 74.6.
$ L& f! f% G) K, a· Compare this value with c2 (n), i.e.with c2 (2016) 6 P6 x, n+ o; e
(c2 is the symbol for theChi-Square distribution)
' h4 c% b" ?* g) ac2 (2016) = 124obtained from c2 table. (For 955 confidence) As n*R2 < c2 ,heteroskedasticity can not be confirmed. * F- ?: ^) r0 o6 F# g1 l
+ M$ A5 U. b" l请参考:regression_explained_SPSS
regression_explained_SPSS.doc
(368 KB, 下载次数: 0)
) I6 _( l, G/ a4 F5 |, Y
|