Research Methods and Statistics

CHAPTER PREVIEW

The simplest experiments involve manipulation of one independent variable to determine the effects on the dependent variable. In this chapter, we expand the possibilities for testing experimental conditions by adding a second independent variable. When we use more than one independent variable, we use an expanded ANOVA model for statistical analysis. Researchers frequently manipulate more than one independent variable in a single experiment. The advantage of a multifactor study (more than one IV) is that we can get a better look at the complex interplay of variables that affect behavior. After all, we are complex, and our behavior is often the result of more than one factor or variable.

In this chapter, we introduce the factorial design. We use two independent variables to illustrate potential interaction effects and we describe how to interpret results of the factorial ANOVA to fully understand the effects of the IVs.

Factorial ANOVA

Expanding on the one-way design, we can manipulate more than one IV in a single experiment. When we have more than one IV, we use a factorial ANOVA for statistical analysis. The advantage to using more than one IV is that you can get a more detailed picture of behavior because you are measuring more complex conditions.

One reason for using multiple IVs is that many psychological questions are too complicated to answer using a single independent variable. Another reason for using more than one IV is that we gain twice as much information from a two-factor study as in a one-factor study. And, even better, we may not need to devote twice the time and energy to get this information because we examine multiple IVs simultaneously.

When we use multiple IVs, each level of one variable is represented at every level of the other variables—the approach is referred to as a factorial design. What do we mean by a factorial (or crossed) design? If there are two variables, one IV with two levels, and a second IV with three levels, we would end up with six different conditions. In other words, each level of the first IV is present in each level of the second IV. The number of conditions, or factors, is derived by multiplying the number of levels of the first IV by the number of levels of the second IV (i.e., 2 * 3 = 6). As illustrated below, three IVs produce a three-dimensional model.