Which of the following is a good rationale for choosing a multilevel design? Group of answer choices The grouping variable is typically divided into two groups. There have been many two-group designs that have established the relationship or effect, and you are examining a new population in which the relationship or effect is not established. Past research has established a difference between two-groups, and adding additional groups is warranted by the research. Research suggests that there may be a linear relationship