One Sample and Paired T-Test
Usage One Sample
Determine if the mean of the values for each node are significantly
different than a specified constant (typically zero).
Usage Paired
The Paired T-Test is also referred to as a T-Test for Dependent Means.
The input metric/shape files must have the same number of columns and
the corresponding columns in the two input files are assumed to be the
same subject. For each node, subtract the corresponding columns
in the input files and determine if their difference is significantly
different than a specified constant (typically zero).
Inputs
- Metric/Shape file containing N columns. Each column
contains data for a single subject and all subjects have performed the
same test.
- (Paired T-Test Only) Second Metric/Shape file
containing N columns. Each column in this metric/shape file
contains data for the same subject as the corresponding column in the
first metric/shape file.
- Fiducial Coordinate file.
- Open Topology file.
- Area Correction metric/shape file.
- Area Correction metric/shape file column number.
- Output Metric/Shape T-Map file.
- Output Metric/Shape permuted T-Map file.
- Paint file containing clusters.
- Text report.
- Number of Iterations. The number of iterations is limited
to pow(2, N).
- Negative Threshold.
- Positive Threshold.
- Alpha (p-value that ranges zero to one).
- Constant to test against T-value (default is zero). This
value should be zero for a Paired T-Test.
Outputs
- Metric/Shape file containing the T-Map computed from the input
Metric/Shape file. This file contains one column.
- Permutation T-Map Metric/Shape file. This file contains
number of iterations columns.
- Paint file identifying clusters.
- Text report.
T-Value Formula
T = [mean - constant] / [sample-std-dev / sqrt(N)]
- mean - average of all values for a node in the input metric/shape
file.
- constant - constant entered by user
- sample-std-dev - standard deviation where the denominator is
(N-1).
- N - number columns (subjects) in the input metric/shape file.
Create Real T-Map
- Using the input metric/shape file, create a new metric/shape file
containing a single column. The value for each node is the
T-Value computed using the input metric/shape file. For a Paired
T-Test, the difference (subtraction) of the two input files
is used to as input to create the real t-map file.
Permutation T-Map
For the number of iterations
- Generate an array containing N elements where each element is a
randomly generated plus or minus one.
- Multiple the elements in the array by the corresponding element
in each row of the input metric/shape file.
- Compute the T-Value for each node.
- Add the T-Value for each node to the output permutation T-Map
file.
Create Output Report
- Find the biggest cluster in each column of the permutation T-Map
metric/shape file and sort them by cluster size.
- Find the largest (alpha)(iterations) clusters in the Permutation
T-Map and use its cluster size as the Significant Cluster Cutoff.
- Find clusters in the Real T-Map file.
- Report all clusters in Real T-Map file that are larger than
Significant Cluster Cutoff.
- Create a paint file showing all significant clusters in Real
T-Map file. Name them plus/minus based upon the threshold they
exceed.