1. 用图表检验Analyze -> regression -> Linear-> Plots
$ U2 r* Z7 D* b4 NScatter plot of the standardised residuals on the standardised predicted values (ZRESID as the Y variable, and ZPRED as the X variable
7 r) s4 c; `' e; _3 B7 n- w0 C' s- A6 X& Q% w
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。
1 h! I5 i4 d7 x$ }% i% ^; L2. 用统计检验
+ b- l; d% {2 D. X- ZHeteroscedasticity——Testing and Correcting in SPSS.pdf
Gwilym Pryce March 2002.doc
(172.5 KB, 下载次数: 4)
! v6 N0 |! d) |4 K3 m* @7 j
- d4 y" F" g' W2 p* L( i9 t; V# ]; ]
Levene’s Test
8 I$ q, Z6 G1 q' M3 ^$ EGoldfeld-Quandt Test! U1 [* s* G) i& Y
Breusch-Pagan Test8 H" t, ~( F) q( B6 s
White‘s Test (比较常用来检验异方差)
& h z( b8 N8 e& u2 j3 R' Z/ KAssume 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)
6 ^# C& Q1 j+ d6 t& o
- u( @' e" _" w1 e) {) b9 g$ T9 k2 E* s! O C o: u
5 d( I u3 Z: o X White’s Test · Calculate n*R2 à R2 = 0.037, n=2016 à Thus, n*R2 = .037*2016 = 74.6.
( Z8 ~$ R a, W$ _9 v; h/ q· Compare this value with c2 (n), i.e.with c2 (2016)
1 D7 s2 w0 C! a+ ^+ A! ~2 |(c2 is the symbol for theChi-Square distribution)
$ {' M4 c& X8 x9 |- z lc2 (2016) = 124obtained from c2 table. (For 955 confidence) As n*R2 < c2 ,heteroskedasticity can not be confirmed. 5 E( k6 z1 L; F- {, P6 [: e! r' N
8 u; C k6 ~; O! M6 z请参考:regression_explained_SPSS
regression_explained_SPSS.doc
(368 KB, 下载次数: 0)
4 |* J- B" V+ e' j z
|