Chapter 13: Factorial Anova in SPSS: Fixed-, Random-, and Mixed-Effects Models

Since its inception, ANOVA has been a popular solution for reducing the experiment-wise error rate problematic in the use of multiple t tests. In the behavioral sciences, factorial ANOVA models are quite popular in terms of assessing mean differences in an experimental or quasi-experimental design. Here we consider the simplest factorial ANOVA model, the two-factor model, where we refer to the factors as Factor A and Factor B. In some cases, factors are known as fixed-effects factors, where the levels of the factor are specifically selected by the researcher. Here only those particular levels of interest are selected for the experiment; results can only be generalized about those selected levels, and thus the same levels would be involved in replications. Some typical fixed-effects factors include specifically selected textbooks for review, socioeconomic level, marital status, age group, gender, or weight.