Ensure all students in the same year group/cohort are tested with the same form (Either Form A or Form B). Question level analysis would not be comparable between the two forms as the questions are different.
Aim to test students within the recommended testing windows. Keep in mind that if you test outside of these windows, the accuracy if the data analysis may be compromised.
Make sure you leave enough time between taking tests of the same level to be able to confidently measure impact of teaching and learning.
If you are going to re-test using the same level, make sure you use the alternate form so that students are not familiar with the questions.