Codigos
de SAS para obtener Levene’s Test y Brown and Forsythe's Test (Tabla 2.3 en Procedimientos Estadísticos para Investigaciones Agrícolas. 1984. Gómez, K.A.; Gómez, A.A. pp. 17-20)
/* Pruebas de homogenidad de varianza*/
data herb;
input tipo $ proporcion
$ DDS $ rendimiento;
cards;
hA/hB21 2.0/0.25 21
3187
hA/hB21 2.0/0.25 21
4610
hA/hB21 2.0/0.25 21
3562
hA/hB21 2.0/0.25 21
3217
hA/hC28 3.0/1.00 28
3390
hA/hC28 3.0/1.00 28
2875
hA/hC28 3.0/1.00 28
2775
hA/hC28 3.0/1.00 28
.
hA/hB14 2.0/0.25 14
2797
hA/hB14 2.0/0.25 14
3001
hA/hB14 2.0/0.25 14
2505
hA/hB14 2.0/0.25 14
3490
hA/hD14 2.0/0.50 14
2832
hA/hD14 2.0/0.50 14
3103
hA/hD14 2.0/0.50 14
3448
hA/hD14 2.0/0.50 14
2255
hA/hE21 3.0/1.50 21
2233
hA/hE21 3.0/1.50 21
2743
hA/hE21 3.0/1.50 21
2727
hA/hE21 3.0/1.50 21
.
hF 1.5 14 2952
hF 1.5 14 2272
hF 1.5 14 2470
hF 1.5 14 .
hA/hB28 2.0/0.25 28
2858
hA/hB28 2.0/0.25 28
2895
hA/hB28 2.0/0.25 28
2458
hA/hB28 2.0/0.25 28
1723
hA/hG28 3.0/1.00 28
2308
hA/hG28 3.0/1.00 28
2335
hA/hG28 3.0/1.00 28
1975
hA/hG28 3.0/1.00 28
.
hA/hD28 2.0/0.50 28
2013
hA/hD28 2.0/0.50 28
1788
hA/hD28 2.0/0.50 28
2248
hA/hD28 2.0/0.50 28
2115
Manual 15-y-35 .
3202
Manual 15-y-36 .
3060
Manual 15-y-37 .
2240
Manual 15-y-38 .
2690
Control - . 1192
Control - . 1652
Control - . 1075
Control - . 1030
;
ods html;
*Levene's test y Brown & Forsythe's test;
proc glm;
class tipo;
model rendimiento =
tipo;
means tipo / hovtest = levene(type=abs) hovtest=bf;
run;quit;
ods html close;
SAS output:
En los
resultados, ambas pruebas no son significativas. (Levene’s test: F10,29
= .88, p = .56; Brown and
Forsythes’s test F10,29 = .53, p =
.85). Entonces, se asume
que existe homogeneidad de varianzas.
No hay comentarios.:
Publicar un comentario