In: Statistics and Probability
We are going over ANOVAs and a question was raised "Let's imagine that we have a research design looking at age of student (traditional versus non-traditional) and mode of learning (campus or online). What would the interaction be? What are the main effects?"
Here we have two factors with two levels each. One factor is the age of students (with levels traditional versus non-traditional) and another factor is the mode of learning (with levels campus or online). Here we can perform two-way ANOVA. Now, suppose we have 2 X 2 table where the rows indicate the age of students and the columns indicate the mode of learning. So, basically, we have four distinct cells. Now, here the main effects are the effects of the age of students and the effects of the mode of learning. Here the interaction effect is the effect due to the interaction of the age and the mode of learning of students. Now, in each cell, if we have only one observation i.e. we don't have any replications in this experiment then we can't estimate the interaction effect and no test can be performed to check the interaction effect. But if we have replications i.e. we have more than one observations in each cell then we can have the estimate of the interaction effect and we can test for it.