Chapter 14, Table 10: One-Within One-Between

(Multivariate Approach) via SPSS Point and Click

The hypothetical data contained in Table 14.10 contains an additional level of angle (four degrees) that was not considered in Table 14.8.

The basic analysis of the data given in Table 14.10 begins by making use of the Repeated Measures procedure. As before, the Repeated Measures procedure can be utilized by first clicking on Analyze, then clicking General Linear Model and then clicking the Repeated Measures tab. Once the Repeated Measures Define Factor(s) menu is displayed, we must define the within-subjects factors. We do this by specifying a name of the factor (we use angle) and the “3” for the number of levels. After we click Add to add the factor we click Define.

One the Repeated Measures box comes up, we specify angle0, angle4, and angle8 as our three within subject factors. This time, however, we also have a between subject factor. The between subject factor, group in this case, is moved to the Between-Subjects Factor(s) box.

Before clicking OK, it is also beneficial to click the Post Hoc button and to move groups to the Post Hoc Tests for box and specify one or more of the Type I error rate adjustment procedures. It may also of interest to click the Options button and then specify that the parameter estimates should be given as well as to have the means displayed for the effects of interest by moving those effects to the Display Means for box.

At this point we have nearly exhausted the commonly used resources available via the SPSS point and click menus. At this point we have not yet replicated many of the results given in the book that make use of Table 14.10. However, most of the other results reported in the book can be obtained by making use of SPSS syntax or by using SAS.

At this point for the present analysis clicking OK will yield the (rather basic) results from the SPSS point and click menus for repeated measures. Note that, as before, for the within-subject effects both the univariate and the multivariate approaches are performed by default. Here we are only concerned with the multivariate approach to repeated measures.