G***G
发帖数: 16778 | 1 y depends on three variables, a, b, and c
so which method is good to determine if there are interaction between a, b,
and
c?
and then how to determine if there are interaction between a and b,
a and c,
and b and c?
which method is best for this purpose?
thanks. | G***G
发帖数: 16778 | 2 please help.
,
【在 G***G 的大作中提到】
: y depends on three variables, a, b, and c
: so which method is good to determine if there are interaction between a, b,
: and
: c?
: and then how to determine if there are interaction between a and b,
: a and c,
: and b and c?
: which method is best for this purpose?
: thanks.
| p**********l
发帖数: 167 | 3 you can model interaction with multiple linear regression in R
for example, y~a+b+c assumes no interaction, and
for interaction, use y~a+b+c+a:b+a:c+b:c+a:b:c
a:b/a:c/b:c are two-way interaction terms
a:b:c is the three-way interaction term
if a/b/c are continuous, t-test in summary() should be enough to see if the
interaction terms are significant. if categorical, use Anova(). | p**********l
发帖数: 167 | 4 also keep in mind, with the presence of interaction terms, you can not
interpret the significance of main effects. |
|
