Visualization and naming convention of output files
All result files are stored in the same folder of the selected SPM.mat file. You can display these results with Show results. Furthermore, the logarithmically scaled p-values are ideal for visualization with cat_vol_slice_overlay for volume files or cat_surf_results for surface and volume files. Here, any statistical thresholds can be easily applied during visualization. Two different corrections for multiple comparisons are available, which are indicated in the file name: family-wise error (FWE) and false discovery rate (FDR).
TFCE statistic
These are the result files of the non-parametric TFCE statistic and also the prefered results, because both voxel-height as well as cluster size are combined.
| TFCE_xxxx | TFCE values for contrast #xxxx |
| TFCE_log_p_xxxx | log-scaled p-values for TFCE-statistic (uncorrected) for contrast #xxxx |
| TFCE_log_pFDR_xxxx | log-scaled p-values for TFCE-statistic (corrected using FDR) for contrast #xxxx |
| TFCE_log_pFWE_xxxx | log-scaled p-values for TFCE-statistic (corrected using FWE) for contrast #xxxx |
T statistic
These are the result files of the non-parametric T-statistics and only available for completeness.
| T_xxxx | T values for contrast #xxxx |
| T_log_p_xxxx | log-scaled p-values for T-statistic (uncorrected) for contrast #xxxx |
| T_log_pFDR_xxxx | log-scaled p-values for T-statistic (corrected using FDR) for contrast #xxxx |
| T_log_pFWE_xxxx | log-scaled p-values for T-statistic (corrected using FWE) for contrast #xxxx |
Null distribution
This is the null distribution based on the permuted design, which might be helpful to evaluate specificity (i.e. false positives) between different approaches.
| TFCE_xxxx | Null distribution for contrast #xxxx |
| nullT_xxxx | Map of T-values (uncorrected) from the permuted design for contrast #xxxx |