statcond

# statcond - statcond Usage Inputs data compare two or more...

This preview shows pages 1–2. Sign up to view the full content.

% statcond() - compare two or more data conditions statistically using % standard parametric or nonparametric permutation-based ANOVA % (1-way or 2-way) or t-test methods. Parametric testing uses % fcdf() from the Matlab Statistical Toolbox. Use of up to % 4-D data matrices speeds processing. % Usage: % >> [stats, df, pvals, surrog] = statcond( data, 'key','val'. .. ); % Inputs: % data = one-or two-dimensional cell array of data matrices. % For nonparametric, permutation-based testing, the % last dimension of the data arrays (which may be of up to % 4 dimensions) is permuted across conditions, either in % a 'paired' fashion (not changing the, e.g., subject or % trial order in the last dimension) or in an umpaired % fashion (not respecting this order). If the number of % elements in the last dimension is not the same across % conditions, the 'paired' option is turned 'off'. Note: % All other dimensions MUST be constant across conditions. % For example, consider a (1,3) cell array of matrices % of size (100,20,x) each holding a (100,20) time/frequency % transform from each of x subjects. Only the last dimension % (here x, the number of subjects) may differ across the % three conditions. % The test used depends on the size of the data array input. % When the data cell array has 2 columns and the data are % paired, a paired t-test is performed; when the data are % unpaired, an unpaired t-test is performed. If 'data' % has only one row (paired or unpaired) and more than 2 % columns, a one-way ANOVA is performed. If the data cell % array contains several rows and columns, and the data is % paired, a two-way repeated measure ANOVA is performed. % NOTE THAT IF THE DATA is unpaired, EEGLAB will use a % balanced 1 or 2 way ANOVA and parametric results might not % be meaningful (bootstrap and permutation should be fine). % % Optional inputs: % 'paired' = ['on'|'off'] pair the data array {default: 'on' unless % the last dimension of data array is of different lengths}. % 'mode' = ['perm'|'bootstrap'|'param'] mode for computing the p-values: % 'param' = parametric testing (standard ANOVA or t-test); % 'perm' = non-parametric testing using surrogate data % 'bootstrap' = non-parametric bootstrap % made by permuting the input data {default: 'param'} % 'naccu' = [integer] Number of surrogate data copies to use in 'perm' % or 'bootstrap' mode estimation (see above) {default: 200}. % 'verbose' = ['on'|'off'] print info on the command line {default: 'on'}. % 'variance' = ['homegenous'|'inhomogenous'] this option is exclusively % for parametric statistics using unpaired t-test. It allows % to compute a more accurate value for the degree of freedom % using the formula for inhomogenous variance (see % ttest2_cell function). Default is 'inhomegenous'. % % Outputs:

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 08/29/2011 for the course CHE 10 taught by Professor Toupadakis during the Spring '08 term at UC Davis.

### Page1 / 10

statcond - statcond Usage Inputs data compare two or more...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online