When testing for the effect of a particular variable on an outcome, it is important for that variable to vary, but only one other relevant variable has to be controlled and the rest can vary.
Items that test for misconception CVM021 in this project (Original Project)
and key idea ()