Hi,
We are using the "evenly present elements" function of the randomizer to assign subjects into different groups uniformly. Also, our study has an attention check and for subjects who failed attention check, we don't want to count their participation. We have to randomize before the attention check as the test depends on the group assignment.
We wonder if there is a way to turn back randomizer counts for failed subjects so that we can still keep evenly presented groups given one group may potentially have more failed attention checks than the others?