Subjects with a diagnosis of ADHD (n = 536), autism spectrum disorder (n = 49), schizophrenia (n = 2) and intellectual disability (n = 2) were also excluded from analysis.  In total, data was available for 9537 subjects. The following filters were applied:

Prodromal Psychosis Score

From "Sum scores on mental health", the PPS sum score was taken as the sum of the number of yes responses (21 questions). (mhy$pps_y_ss_number).
A severity score was taken as the sum of scores for how much each symptom was bothersome. (mhy$pps_y_ss_severity_score).
These scores were taken as continuous independent predictor variables in the regression model. Separately, these scores were binarised to taken in to account the top 170, 199 respectively (cut-off for SUM yes scores binarised at PPS = 11; cut-off for binarised severity scores = 31). 

fMRI Activity: Monetary Incentive Delay

The beta weights for MID anticipation of reward vs. neutral contrast were taken as the dependent variable. Remove subjects for whom participant did not have an acceptable performance in task (this was assessed based on the threshold that all trial types must yield more than 3 events for both positive and negative feedback). No. excluded approx. 244 (after other things excluded like ADHD etc. as above) (mid$beh_mid_perform_flag). Final numbers = 6658.
These were further filtered to remove subjects with MID values above or below 3 standard deviations.

Covariates

Motion: Mean framewise motion for scans was included as a covariate (midaparc01$fmri_beta_gparc_mean_motion)
Site: In order to control for variance due to site, we included the variable "device serial number" which indicates the individual MR scanners used to acquire data. In this dataset there were 27 separate devices included, however two had no data leaving 25 separate scanners. The number of subjects acquired per scanner varied from a minimum of 21 to a maximum of 816 (see Figure \ref{524027}).