08_methods_example.tex

We apply our estimation procedure to data from a study of 13 subjects on the role of higher order visual areas in imagined visual motion \citep{Seurinck2011}.  First level analyses were carried out with a standard General Linear Model approach using FSL's FEAT function \citep{Jenkinson2012}.  The second level analysis is performed in R with OLS, resulting in a 3D $T$-statistic map, that is transformed to $Z$-values \citep{Hughett2007}.

For this one Z map we apply our method with $u=2.5$, characterizing the signal with $\hat\pi_1$, $\hat\mu_1$, $\hat\sigma_1$, and computing prospective power for different statistical thresholds (uncorrected, FDR-corrected, FWER-corrected with a Bonferroni procedure and FWER-corrected with a Random Field Theory procedure) and sample sizes ($n^*=13,...,100$).