TOSTER icon indicating copy to clipboard operation
TOSTER copied to clipboard

Equivalence test for interactions

Open Lakens opened this issue 6 years ago • 0 comments

In the past I asked you about how to run equivalence test for interaction effects. I talked to my stats professor here at Cornell and he suggested using emmeans in R. Here is the code (though it does not create nice plots like your package):

categorical interaction equivalence test

summary(contrast(emmeans(lm1, c("x1", "x2"), weights = "equal"), interaction = "consec" ), infer = TRUE, null=0, side="equivalence", delta=.1)

continuous interaction equivalence test

r1 <- ref_grid(lm1, at = list(x1 = c(0, 1), x2 = c(0, 1))) summary(emmeans(r1,c("x1","x2"),contr = list(int=c(1,-1,-1,1))),infer=TRUE,null=0,delta=.1,side="equivalence")

Lakens avatar Aug 01 '19 14:08 Lakens