Figure 4 shows numerical simulation of the expected -score increase for different first-level variance configurations. As before, the expected increase is independent of the second-level effect size but will depend on the first- level variance configuration for group 1 and group 2 as well as the different second-level variances and . The added flexibility of the heteroscedastic model is important for a variety of real FMRI experiment where the two groups naturally will have different variance configurations, e.g. studies of patients vs. non-patients. Once again, significant changes in -score (e.g. ) can be seen over a large set of configurations.
|