Prism can perform three-way ANOVA with only certain kinds of data:

•Only two levels of two of the factors. For example one factor could be male vs. female, the other factor control vs. treated, and the third factor could be three different time points. But Prism can only handle one factor with more than two levels. For example, Prism could not deal with the example above if the second factor had a third level: control vs. treated vs. treated + antagonist.  

•No repeated measures. A repeated measures design would be, for example, when the same subject is measured at several times. Prism cannot handle repeated measures with three-way ANOVA (but it can with one- or two-way ANOVA). Please write us if you have data best analyzed with three way ANOVA with repeated measures in one factor (or two or three factors) with the details.

•No missing values. The number of replicate measurements (or subjects) for each of the combinations of treatments  must be identical. Statisticians would say the design must be "balanced". Prism cannot yet handle missing values (an unbalanced design) in three way ANOVA (as it can for one- and two-way ANOVA).

•You must enter two or more values for each combination of conditions (alternatively, enter the mean, sample size and SD or SEM). Prism cannot compute three-way ANOVA with only a single value for each set of conditions.