Comparing more than two means: one factor ANOVA with independent samples. Multiple comparison testing with the Newman-Keuls test - Improving the User Experience through Practical Data Analytics - page 147

SIDEBAR: A MINI-GLOSSARY OF RELEVANT TERMS THAT YOU NEED TO
KNOW!—cont'd

it's a virtual guarantee that you will have replication. For example, you have multiple partici-
pants, each providing his/her opinion about the sophistication of different designs; you have
replication. In this chapter, we sensibly have more than one person from each age group evaluat-
ing the sophistication of designs, and as a consequence we have replication.
However, when there are two or more factors, it is not automatic that you have replication
(if you do not have replication, we say you have “no replication”). In Chapter 7, we discuss a
two-factor design with no replication. In Chapter 8, we discuss a two-factor design which does
have replication. In these chapters, we discuss the role of replication, and again, why it may
matter when conducting your analysis!

• Interaction : When there are two or more factors under study, the factors may have “interaction”

between them. The idea of interaction is similar to the notion of “synergy.”
If there is no interaction between the (say, two) factors, then each factor is separately having
whatever effect it has, if any, on the dependent variable and the effects of the two factors are additive .
However, if the effect of having each of the two factors at a speciic level is not simply additive
(the sum of the individual effects), then we say that there is interaction (or synergy) between them.
Whether there is interaction between factors or not can be extremely important, as you will see in
Chapters 7 and 8.

FIGURE 6.4
Data in SPSS format.