Statistical methods in EMEGS comprise cross-correlation,
custom within contrasts, t-tests and repeated measures ANOVA. Each will
be described seperately in the following.
Cross-correlation
T-Tests
T-Test are available from the Emegs2d
'Calculate\T-Test\'-Menu. You can choose between one sample, two sample
and paired t-test. Special cases of these are the restricted paired
t-test, which defines a minimum criterium for significant points to be
displayed in terms of a minimum of significant neighbour channels and
a minimum number of significant subsequent
time points. Symmetry tests (needs to be written)...
For all of these tests, you need a textfile, listing the paths of the
average SCADS files, that you wish to use for calculation.
Contrasts
Contrasts are very similar to T-Tests, except
that they allow the testing of multiple means with a set of user
defined coefficients. In EMEGS, contrasts are always within contrasts,
therefore the subject order in the batchfile must be consistent across
all conditions. If you choose 'Contrasts from list', you are given a
set of predefined coefficients, of which you can choose. Otherwise, you
have to enter the coefficients yourself. In either case, the order of
your coefficients must correspond the condition order in your batchfile
and their total must equal zero. Moreover, you can choose to use a
normal distribution for the testing, or to use a distribution,
generated by ..... (permutation test).
For instance, consider a contrast to compare 3 groups of visual evoked
potentials: those evoked by pleasant pictures, those evoked by
unpleasant ones and those evoked by neutral ones. And suppose your
hypothesis would be that over visual areas, the potential for emotional
pictures is more negative than that of neutral pictures. To test this,
you would make a batchfile with the path of all pleasant average files
(subject 1 to say 16), followed by the paths of all
unpleasant ones (subject 1 to 16), followed by the neutral ones in the
same order. Your coefficients would be : -1 -1 2, indicating that
both the unpleasant ones and the pleasant ones are more negative than
the neutral ones.
Repeated Measures ANOVA
EEG and MEG
studies are most often analyzed with a special kind of analysis of
variance, that accounts for within subject changes rather than for
between subject differences. EMEGS offers the possibility to calculate
this kind of analysis directly, without exporting the data to a
separate statistic software package. Moreover, you can choose between a
region-of-interest analysis, averaging over sensorgroups and time
points, or a complete analysis for every sensor and time point in your
data.
To run a repeated measures ANOVA, prepare your data as described:
<><>Each condition for every subject has to be saved as an SCADS average file. Every file has to have the same number of points and number of channels and the same baseline calculation. You need a textfile, listing the paths of those files on your machine (a 'batchfile'), with one path per line.This batchfile has to reflect the design of the planned analysis, that is, your paths have to be listed according to the hierarchy of your factors. The lowest level is always the subject factor, so you 'll start with one cell for which all subjects average files are given. Beneath that, you list all subjects average files for the next cell etc. The subject order has to be identical in every cell, and you have to have equal number of subjects in every cell. Unequal cell sizes are not yet supported. Cells are ordered from lowest hierarchy position of the factor to highest hierarchy position. For instance, consider a 2X2 design with the factors 'task' (count forward vs count backward) and 'color' (count red squares vs. count green squares). Your batchfile for 16 subjects has to have the following form:
Once you've prepared this, load one of the listed files in
Emegs2d, set the baseline and display all points. Start Emegs3d. Here,
load the batchfile as filematrix (File\FileMatrix\LoadBatchfile).
Then choose \Calculate\Repeated Measures Anova\Define enter enter your
design. You always have to enter the number of subjects in your sample
(ignoring betweenfactors). If you're calculating the design for all
points and channels set nrofintervalls to the total number of points
in your data files and nrofchannelgroups to the number of channels
(matlab 6 and lower) or check the 'all points and channels' checkbutton
(matlab 7). If you want
to run one special analysis across a selected number of channels and
certain time points, enter the number of channelgroups and the number
of timeintervalls correspondingly. These values have to match the
channelgroups loaded in Emegs2d and the time/intervall settings in
Emegs3d. If more than one of each is provided, channelgroup and/or time
will be a separat factor in the ANOVA. Please not that for a pointwise
analysis, the whole intervall has to be selected in Emegs3d including
the baseline points!!!
Click 'OK' and choose \Calculate\Repeated Measures Anova\Run Anova. For
a pointwise analysis, you will be prompted for a target folder, where
results are going to be saved. EMEGS will save two average file in
SCADS
format for every factor and interation in your ANOVA, one with
p-values,
one with the F-Values. For a single analysis results will be
displayed in new figure, from which you can calculate post-hoc test,
display means graphically and export the data to a text file.