# Multivariate analysis of BOLD activation patterns recovers graded depth representations in human visual and parietal cortex
We presented subjects with stereoscopic sphere stimuli, composed of multicolored flickering dots. Stimuli were presented at various positions evenly spaced along the horizontal (left-right) spatial axis, and the depth (near-far) spatial axis. We used an inverted encoding model to reconstruct the locations of stimuli along each of these spatial axes.
For analysis code, see our Github repository at: https://github.com/mmhenderson/IEMdepth
#### Contents:
Each file in IEMdepth_trialData contains all data for a single subject.
The variable "trialDataAll" is a structure array containing data for 12 ROIs. The names of the ROIs are given in "VOIs".
For each element in trialDataAll, the fields are as follows:
- aAll [nTrials x nVoxels]: the average signal in each TR on each trial, obtained by averaging the 3rd and 4th TRs following stimulus onset. This was used for all analyses reported in our paper.
- bAll [nTrials x nVoxels]: beta weight for each voxel on each trial, obtained with a GLM. These values were not used for any of our main analyses.
- stimLocsAll [nTrials x 3]: the exact [x,y,z] coordinates of the sphere stimulus presented on each trial, in OpenGL coordinates. Note that the second column, Y position, is always 0.
- condAll [nTrials x 1]: the task condition corresponding to each trial, where 1 is attend fixation and 2 is attend target. We did not analyze the attend target data in our paper.
- runsAll [nTrials x 1]: the run labels for each trial. There are 36 trials per run, with runs always occuring in task pairs - e.g. run 1 of the fixation task, followed by run 1 of the target task.
- spinDir [nTrials x 1]: the direction of rotation of a brief pulse of motion by the sphere stimulus, which occured on every trial (this was relevant only on attend target trials).
- fixChgDir [nTrials x 1]: the change in contrast of the fixation point, which occured on every trial (this was relevant during attend fixation trials).
- resp [nTrials x 1]: the subject's response button on each trial. A correct response should match the value in spinDir during attend target runs and should match the value in fixChgDir during attend fixation runs.
