1. 用图表检验Analyze -> regression -> Linear-> Plots9 B" i* s9 s8 m. T9 a
Scatter plot of the standardised residuals on the standardised predicted values (ZRESID as the Y variable, and ZPRED as the X variable- K0 A: N) S" z' O7 H/ Z
6 Y4 ^' z9 _, d7 u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。; z8 w% ]9 j+ ` j$ k+ ^' V
2. 用统计检验; o1 q4 {. x1 E$ u$ Z
Heteroscedasticity——Testing and Correcting in SPSS.pdf
Gwilym Pryce March 2002.doc
(172.5 KB, 下载次数: 4)
/ {" _; D0 x g; I
. p+ ]0 `3 S1 \2 H* r# l, C
8 z+ k! u0 f" S1 A3 G4 HLevene’s Test
5 ~# ?6 i8 t; ?$ w5 h3 uGoldfeld-Quandt Test1 E+ t* h) s, Z5 T, Z- H- p, L
Breusch-Pagan Test
" m5 E7 l$ A- M tWhite‘s Test （比较常用来检验异方差）
3 s; ~3 d! J3 F3 U: W/ PAssume 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)
3 z9 i. l2 D" H
5 `$ [6 s; |- D8 D4 [/ k W$ p# m- ^, ^3 l: I9 L3 \
+ B' q0 O1 {* L: ?! d: n White’s Test · Calculate n*R2 à R2 = 0.037, n=2016 à Thus, n*R2 = .037*2016 = 74.6. ( T+ t# a }7 r! W
· Compare this value with c2 (n), i.e.with c2 (2016) 2 i5 l0 }: m- W; N+ ~
(c2 is the symbol for theChi-Square distribution) ' z5 r& s' B; v
c2 (2016) = 124obtained from c2 table. (For 955 confidence) As n*R2 < c2 ,heteroskedasticity can not be confirmed. ' N# [6 B3 B8 K( c8 b3 I
% y6 e6 B) G8 D; X4 T; }
请参考：regression_explained_SPSS
regression_explained_SPSS.doc
(368 KB, 下载次数: 0)
; z3 H1 F8 p, N6 X" O, d$ n% g
| 
IP卡
狗仔卡
发表于 2013-10-13 16:12
转播
淘帖
分享
收藏
支持
反对
微信
提升卡
置顶卡
沉默卡
喧嚣卡
变色卡
抢沙发
显身卡
