ABSTRACT
This chapter outlines the development of the mixed-model approach. It considers two-way factorial designs where both factors are within-subjects. The chapter reviews the two-way designs where one factor is between-subjects and the other is within-subjects. It compares the multivariate and mixed-model approaches for these two types of designs. The advantages and disadvantages of the multivariate and mixed-model approaches in the split-plot design are essentially the same as in completely within-subjects designs. The analysis of data from higher-order within-subjects designs can obviously become quite complicated. Although the multivariate approach does not require the sphericity assumption of the mixed-model approach, the multivariate approach nevertheless shares several assumptions in common with the mixed-model approach. The term in the first set of parentheses is the quadratic trend for angle when noise is present, whereas the term in the second set of parentheses is the quadratic trend for angle when noise is absent.
