During the nonmatching-to-place task, MD lesion rats performed a small percentage of correct choices during the four blocks of acquisition and showed a very strong perseverative behavior.
A separate analysis considered whether the first reference memory error made in each testing trial was more likely to be to the single, never-baited arm or one of the adjacent, never-baited arms, over each of the eight, four-trial blocks of acquisition.
Late Acquisition was defined as block 3 of Acquisition, when discrimination learning was at maximum; significant interactions were followed up by univariate analyses of covariance (ANCOVAs), with demographics, childhood trauma (CTQ), and adult trauma (TEI) used as covariates in all analyses involving diagnostic groups.
360 T2-weighted echo-planar images depicting BOLD contrast were acquired per block of scanning acquisition, with the first 10 discarded to avoid T1 equilibrium effects.
We measured the percentage of correct responses for both groups during the last block of the acquisition period and during the first block of the memory task performed 30 days later (Figure 6A C).
Within each block of scanning acquisition, volunteers monitored 22 targets, 22 morphs from each of the three degrees of similarity, 22 same type distractors, and 74 other type distractors.
In the group level analysis, each volunteer contributed six whole brain images, containing the parameter estimates for the similarity and ambiguity regressors, separately for each block of scanning acquisition.
To study savings in particular, we compared the percentage of CRs between the first block of the acquisition experiment and the first block of the savings experiment.
The first ANOVA examined the effects of similarity to the current target object averaged over the three blocks of experimental acquisition (Figure 5).
