Description of data for: "Emerging Representational Geometries in the Visual System Predict Reaction Times for Object Categorization" by J. Brendan Ritchie, David A. Tovar, and Thomas A. Carlson ------------------------------------------------------------------------------------ Contents: Data consists of a folder containing 29 "S1" data files (Matlab .mat files), one for each of the subjects whose data was used in our experiment. Each file has two structure: "ExptLabels" and "Results". 1. "ExptLabels.var" is a 1x4 cell containing structures with information regarding the labels and categories of the image stimuli used in the study. 2. "Results" is a structure containing results from performing multi-variate pattern analysis on data collected using magnetoencephalography (MEG), as well as descriptive statistics of reaction time data. Each “Results” structure contains the following: 2.1 "Results.EventLength" states the period our MEG data was epoched. 2.2 "Results.ActivePassive" is a 1 x 2 structure with four fields. The fields in Row 1 contain data collected while subjects performed the “active” categorization task from our experiment. The fields in Row 2 contain data collected while subjects performed the “passive” distracted viewing task from our experiment. Three of the fields are of interest: 2.2.1 "Results.ActivePassive.fullcross_pcorr" is a 1 x 36 vector containing the % correct classifier performance for linear classifiers trained and tested at each of the 36 time points (20 ms resolution) of the MEG data. 2.2.2 “Results.ActivePassive.fullcross_dpr” is a 1 x 36 vector containing the d’ value for classifier performance for linear classifiers trained and tested at each of the 36 time points (20 ms resolution) of the MEG data. 2.2.3 “Results.ActivePassive.RTdistanceMatrix” is a 36 x 24 x 3 matrix. (:,:,1) of the matrix contains the mean RTs for each of the 24 stimuli, repeated 36 times (once for each time point). (:,:,2) of the matrix contains the median RTs for each of the 24 stimuli repeated 36 times. (:,:,3) contains the Euclidean distances for each of the 24 exemplar stimuli at each of the 36 time points. These are distances of a pattern of activity at a time point from a hyperplace along a discriminant axis, which was computed using linear discriminat analysis. ----------------------------------------------------------------------------------- Plotting 2.2.2 recreates Figure 3 from out manuscript. Measuring the rank order correlations between the distances and RTs, using the data in 2.2.3, generates the findings depicted in Figures 4-6, and Figure S1, of our manuscript.