| Cognitive processes can be studied with functional magnetic resonance imaging (fMRI) experiments. Different within subject and between subject designs exist with their own advantages and disadvantages. This research project aims at finding optimal designs for fMRI experiments that have maximal efficiency and maximal power for finding real effects. By means of results from the statistical theory of optimal designs for generalized linear mixed effects models, including both random and fixed parameters together with (auto)correlated errors, the problem of finding optimal designs can be formulated as a nonlinear optimisation problem. The optimal designs will be empirically evaluated with real fMRI data. |
