ANCOVA was conducted on a per-voxel basis, and the resultant β maps were thresholded in the manner described above, with the exception that F-tests were used. Because testing laterality determines whether the β in a voxel in the right hemisphere is significantly different from the β in the homologous voxel in the left hemisphere, half as many tests were conducted as in a nonlaterality analysis. Therefore, a mask containing only the right-hemisphere portion of the superior prefrontal mask was used. Anxious apprehension ROI mediation analyses Mediation analyses Inhibitors,research,lifescience,medical were carried out in SPSS v19 using the INDIRECT macro (Preacher and Hayes 2008). PSWQ was entered as an independent variable, with MASQ-AA and MASQ-AD-LI entered as Inhibitors,research,lifescience,medical covariates. To isolate behavioral habituation to negative stimuli, composite RT and error variables were created. Specifically, interference due to negative words (i.e., negative – neutral) in the first half of the task was subtracted from interference due to negative words in the second half
of the task. Mediators were average habituation in activation to negative words in each ROI associated with PSWQ. Tests of specificity to negative stimuli Inhibitors,research,lifescience,medical Examination of positive stimuli In order to ensure that present findings were driven by the negative valence of the stimuli rather than their arousal value, the relationship between PSWQ/MASQ-AA and habituation of activation related to positive words was examined in two ways. Inhibitors,research,lifescience,medical First, the analyses above were rerun with the exception that the single-subject contrast was positive minus neutral (as opposed to negative minus neutral) and that 2-tailed tests were used. Apart from this difference, these analyses were identical to the main analyses. Second, for each ROI identified in the main (i.e., negative minus neutral) analyses, the average β (across voxels) was computed for positive and neutral (vs. baseline), for each half of the session, for each participant. These values were entered into a repeated-measures GLM (using SPSS v19), with Time (first half of the task vs. Inhibitors,research,lifescience,medical second half of the task) and Emotion
(positive vs. neutral) Pharmacological Reviews as the repeated factors, and PSWQ, MASQ-AA, and MASQ-AD-LI as continuous predictors. Of specific interest were the Time × Emotion × PSWQ and Time × Emotion × MASQ-AA interactions (depending on whether the ROI was associated with PSWQ or MASQ-AA). For brevity, only findings for these effects are reported. These analyses are only partially independent (Kriegeskorte et al. 2009), because the neutral condition was part of the contrast used to define the ROIs. However, the lack of complete independence biases toward finding patterns similar to those observed in the main analyses and thus actually biases against the test of specificity. Therefore, this bias click here renders the tests more conservative for present purposes.