Cohen.d gives wrong value when data is not arranged by f

[juyujeng]

when using cohen.d(d ~ f, data)
if data is not arranged by the f, it might give a wrong value of Cohen'd

[mtorchiano]

I'm, not sure I understand you issue, could you please post a minimal working example?

[juyujeng]

Sorry I didn't make my question more clear.
Here is my exmaple scripts.

I found that this problem happens when unequal sample sizes of the two groups.

library(effsize)
d.data <- data.frame(group = sample(c(1,2), 60, replace = T, prob = c(.33,.67)),
                     value = rnorm(60,100,15))
group1 <- d.data$value[d.data$group==1]
group2 <- d.data$value[d.data$group==2]
# the results of these two line bellow are different
cohen.d(value ~ group, d.data)
cohen.d(group1, group2)

d.data.arranged <- d.data[order(d.data$group),]
# after arranged, the results of these two lines bellow are the same
cohen.d(value ~ group, d.data.arranged)
cohen.d(group1, group2)

[mtorchiano]

Thanks, I'll try to have a look at it as soon as I have some spare time

[mtorchiano]

Issue resolved!