Let’s use SPSS to conduct a two-way ANOVA for the data on myth busting that we used in this chapter. We enter the data in three columns—one for each participant’s scores on each independent variable (age and number of repetitions) and one for each participant’s score on the dependent variable (false memory).

We instruct SPSS to conduct the ANOVA by selecting: Analyze → General Linear Model → Univariate and selecting the variables. (Even though we have two independent variables, we still choose univariate, which means one variable, instead of multivariate, which means many variables. That’s because this choice refers to the dependent variable rather than the independent variable, and there’s just one dependent variable, false memory.)

Select the dependent variable, false memory, by highlighting it and clicking the arrow next to “Dependent Variable.” Select the independent variables (called “fixed factors” here), age and repetitions, by clicking each of them, then clicking the arrow next to “Fixed Factor(s).” To include specific descriptive statistics, as well as a measure of effect size, select “Options,” then select “Descriptive statistics” and “Estimates of effect size.” Then, select “Continue.” Finally, click “OK” to run the analysis.

The output in the screenshot shown here includes the same statistics that we calculated earlier. The small differences are due only to rounding decisions. For example, we see that the F statistic for the main effect of age is 13.333. Its p value is found in the column headed “Sig.” and is .006. This is well below the typical p level of 0.05, which tells us that this is a statistically significant effect. The effect size is found in the final column, “Partial Eta Squared,” which can be interpreted as we learned to interpret R². The effect size of .625 indicates that this is a very large effect.

354