# [FieldTrip] freqstatistics ivar and uvar

Irina Simanova irina.simanova at mpi.nl
Wed Jul 11 17:20:41 CEST 2012

```Dear all,

I have a very basic question on freqstatistics. I want to compare the
TFR's in 3 experimental conditions over 12 subjects.  I run the
freqstatistics with cfg.method = 'depsamplesF'.

cfg.uvar = 1
cfg.ivar = 2
cfg.design = [1:12 1:12 1:12; ones(1,12) 2*ones(1,12) 3*ones(1,12)]

I am confused with the the following lines that FT prints out:

repeated measurement in variable 1 over 12 levels
number of repeated measurements in each level is 3 3 3 3 3 3 3 3 3 3 3 3

I assume it is correct, since it's exactly the same output as I get
with the tutorial within-subjects statistics example. But why is not
it the other way around? There are 3 levels of the independent
variable, and the number of samples in each level is 12?

Thank you!

Best,
Irina

Here is the code:
cfg = [];
cfg.latency          = [0 1.5];
cfg.channel = 'all';
cfg.method           = 'montecarlo';
cfg.statistic        = 'depsamplesF';
cfg.correctm         = 'cluster';
cfg.clusteralpha     = 0.05;
cfg.clusterstatistic = 'maxsum';
cfg.tail             = 1;
cfg.clustertail      = 1;
cfg.alpha            = 0.05;
cfg.numrandomization = 100;
cfg.neighbours       = neighbours;
cfg.parameter = 'powspctrm';
cfg.design   = design;
cfg.uvar     = 1;
cfg.ivar     = 2;

[stat] = ft_freqstatistics(cfg, TFRgrand_1, TFRgrand_2, TFRgrand_3);

computing statistic over the frequency range [3.962 39.937]
computing statistic over the time range [0.000 1.500]
selection powspctrm along dimension 2
selection powspctrm along dimension 3
selection powspctrm along dimension 4
using "ft_statistics_montecarlo" for the statistical testing
using "statfun_depsamplesF" for the single-sample statistics
constructing randomized design
total number of measurements     = 36
total number of variables        = 2
number of independent variables  = 1
number of unit variables         = 1
number of within-cell variables  = 0
number of control variables      = 0
using a permutation resampling approach
repeated measurement in variable 1 over 12 levels
number of repeated measurements in each level is 3 3 3 3 3 3 3 3 3 3 3 3
computing a parametric threshold for clustering
computing statistic

```