During the test, participants were shown 36 photographs of eye gazes in a consecutive sequence, and they were asked to pick one term from four possible descriptions of the person whose eyes were portrayed in the photo (for example, anxious, thoughtful, skeptical, suspicious). Behavioral analyses were performed using Matlab statistical toolbook and SPSS. Ordered logistic regression was implemented using the PLUM (polytomous universal model) procedure in SPSS (DeCarlo, 2003). The dependent

variables were the participants’ choices coded as trinary variables (i.e., buy, sell, or stay), while the two dependent measures were market prices (average of best bid and best ask available in the Afatinib mw choice period) and fundamental asset value for the current period ($0.24 × [15 − t + 1]) (dashed line in Figures 1C and 1D). For each model, we reported the Nagelkerke pseudo R2 (Nagelkerke, 1991) and the BIC (Kass and Raftery, 1995).

Forty-five slices were acquired on a 3T Siemens Trio at a resolution of 3 mm × 3 mm × 3 mm, providing whole-brain coverage. A single-shot echo planar imaging (EPI) pulse sequence was used (TR = 2800 ms, TE = 30 ms, FOV = 100 mm, flip angle = 80°). The images were collected at a tilted angle of 30° from the anterior commissure. For each subject, at the end of the first scanning day (day 1), the EPI functional scanning was followed by a whole-brain, high-resolution, T1-weighted anatomical structural scan and local field maps. Image analysis was performed using SPM8 (http://www.fil.ion.ucl.ac.uk/spm/). KPT 330 The first five volumes from each session were discarded to allow for T1 equilibration. Raw functional, structural, and field map files were reconstructed using TBR. Field maps were reconstructed into a single-phase file. This field map file was then used to realign and unwarp EPI functional images. Structural images were reregistered to mean EPI images and segmented into gray and white matter. These segmentation parameters were then used to normalize Metalloexopeptidase and bias correct the functional images. Normalized images were smoothed using an 8 mm full-width Gaussian kernel at half-maximum (FWHM).

A GLM was constructed in which onset regressors (beginning at the start of each video) for each session were assembled by convolving δ functions with a canonical hemodynamic response function (HRF). These regressors were modulated by a parametric regressor coding for the CPV, a combination of the value in cash and in shares held by a subject at each point in time (CPV = cash + [shares × fundamental value at time t]). A correction for temporal autocorrelation in the data (AR 1 + white noise) was applied. Finally, six motion parameters were included in the GLM. In order to find an interaction of the increased value representation due to the bubble manipulation, we contrasted linear increase to CPV in the bubble markets versus the nonbubble markets.