A 2 x 2 factorial design has 2 factors, each with 2 levels, creating a
total of 4 conditions. So you can have factor A with levels A1 and A2,
and factor B with levels B1 and B2, resulting in the conditions A1-B1,
A1-B2, A2-B1, and A2-B2.
Example: Let's say I wanted to know how drinking coffee and eating
breakfast affected exam performance. One factor would be "coffee" and
could have two levels: "drank coffee" or "did not drink coffee". The
other factor would be "breakfast" and could have two levels: "ate
breakfast" or "did not eat breakfast". One of the four possible
conditions is: "did not drink coffee AND ate breakfast".
What you can find using this design are main effects of each factor
and whether there is an interaction. So there might be a main effect
of drinking coffee on exam performance. There might also be a main
effect of eating breakfast on exam performance. But perhaps, when
someone drank coffee AND ate breakfast, there was a greater effect
than either factor by itself. That is known as an interaction. |
