To run program: valmap input_file output_file Parameters for input files (so far): Number of dependent variables, use -(number) is dependent variable is NOT spatially varying Number of spatially varying independent variables number of scans (or observations) number of effects number of confounds normalize dependent variables (needed if longitudinal atrophy rate maps are uncorrected for interval) number of contrasts number of permutations separate files (1=dependent variables in separate files, 0=dependent variables in separate frames, 0 for univariate) mask_images (1=images are masked with mask file provide, 0=no masking) basename1 (name for files containing beta estimates) basename2 (name for files containing beta estimates for reduced model) statname (name for files containing F or T statistics) maskname (name of mask file...if mask_images==0 this isn't used but some string is expected. Sorry!) Design matrix--IV intercept DV1 DV2 DV3 normval (note: spatially varying DV must be placed last but before normval) list contrasts, one contrast per line (e.g., 1 0 0 0) NOTE: if normalize dependent variables == 1, then a normalization factor is expected for each scan (or observation), and must be entered with the design matrix as *the last column*. I have used this normalization factor when analyzing jacobian maps of longitudinal atrophy to account for differences in interscan interval (thus, each jacobian map was divided by interval, normalizing atrophy to rate/yr). If analyzing perfusion maps, one might want to normalize by perfusion within some region of interest. CAUTION: if doing multivariate analysis, this will normalize all subject's dependent variable images by the same normval, and that's probably not what you want to do (in that case, you'll have to normalize the images yourself some way before statistical analysis). ALSO: normval will only normalize the dependent variable images. If you need to normalize your independent variable images, you'll have to do that before statistical analysis also. Output Files: basename1X.nii--regression coefficients, separate file for each independent variable basename2X.nii--regression coefficient for reduced model (intercept only) statname.nii--F statistic for overall model (this is F for ANOVA) statnameX.nii--t-statistic for each independent variable statnameR2.nii--Rsquared for overall model (can compare these for model fit) basenameEstContrastX.nii--size of effect for each specified contrast statnameFContrastX.nii--F statistic for each specified contrast If multivariate analysis (i.e., more than one dependent variable), then regression coefficient and t-statistic files will be multi-frame, with each frame showing results for a single independent variable (e.g., frame 1 will show group regression coefficients for group on dependent variable 1, frame 2 will show group regression coefficients for group on dependent variable 2, etc). Multivariate analyses will have additional output files: statnameIVOverall.nii--F statistic for overall effect of each independent variable on all dependent variables; this will be a multi-frame file, with each frame showing overall F-statistics for a single independent variable (e.g., frame 1 will show overall effect of group on all dependent variables, frame 2 will show overall effect of age on all dependent variables, etc.) statnameIVContrastOverall.nii--F statistic for overall effect of each contrast on all dependent variables; this will be a multi-frame file, with each frame showing overall F-statistics for a single contrast.