Check 'Parameter estimates' so that we will see the true mean differences.If you would like to carry out post-hoc comparison, click Post Hoc to specify which factor you would like to look at, and which method you would like to use.The default 'Full factorial' is good, make sure Type III sum of squares was selected.Click and drag the binary and categorical independent variable, x1 and x2, into the 'Fixed Factor(s)' panel.Click and drag the continuous independent variable, y, into the 'Dependent Variable' panel.Analyze > General Linear Model > Univariate.We will be using the continuous variable as the dependent variable, and the three-level categorical variable as the independent variables. Sample file is based on TwoWayANOVA, which is a simulated data with 120 cases and three variables: one continuous, one binary, and one categorical with three levels. General ReferencesĪnalysis of variance on Wikipedia Data Source We will discuss how to incorporate interaction terms. The null hypothesis of a factorial ANOVA is that the population means of all groups are the same.Ī extra complication of factorial ANOVA compared to one-way ANOVA is that we would need to account for interactions between groups using interaction terms. The dependent variable has to be continuous and the independent variables (i.e. The simplest factorial ANOVA is a 2-way ANOVA, which includes two independent categorical independent variables, also referred to as factors. Factorial ANOVA (ANalysis Of VAriance) allows us to compare means of groups across more than one independent variable.

Unianova spss code