Random effects analyses were used for statistical inference on a voxel-by-voxel basis (height threshold p < 0.001, uncorrected). Paired t tests were used to test for significant differences between the odorant classes (e.g., the hypothesis [AO > NApO]). We focused principally on differences between odorant classes, rather than the effects of an odorant class compared to an odorless baseline, so as to control for the nonspecific effects of olfactory stimulation, per se (Kareken et al., 2004). In addition to the voxel-by-voxel analysis, activation data for the effects of interest from a priori volumes of interest in the left and right nucleus accumbens (1536 mm3 volume on each side) were extracted using the Marsbar toolbox for SPM2 (Brett et al., 2002).
