<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none"><!-- p { margin-top: 0px; margin-bottom: 0px; }--></style>
</head>
<body dir="ltr" style="font-size:12pt;color:#000000;background-color:#FFFFFF;font-family:Calibri,Arial,Helvetica,sans-serif;">
<p>Dear Drs. Oostenveld, Maris and fieldtrip community. <br>
I am facing some problems with the statistical analysis of my ERP experiment. <br>
I conducted a ERP experiment using visual stimuli which is characterized by a within subjects design. I have 8 independent variables (4 active stimuli and 4 control stimuli) and since both the paradigm and the stimulus set used are novel I am not making a-priori
assumptions about the effects of interest in time and space. <br>
In order to analyse my data I am planning to perform a cluster based permutation test based on a repeated measures factorial Manova.
<br>
My aim with this test is to: find the F value associated to each of my conditions at each time point/electrode. Find clusters of significant F values in time and space and then perform the cluster-based permutation test with the appropriate non-parametric correction.
</p>
<p><br>
Given the fact that I am new to fieldtrip I would like to know if such analysis can be performed using the option: 'ft_statfun_depsamplesFmultivariate'. If it is possible, should I specify my design only by appropriately pairing subjects/condition in the 2
row design vector? I have 3 factors with 2,3,3 levels respectively and I am not sure I am properly conducting the analysis.
</p>
<p> </p>
<p>Finally does this approach make sense or it is unnecessarily complicated? I was previously thinking to do this type of analysis using depsampleT statistics, however I will have to perform 16 cluster-based permutation t-tests in order to test my effects of
interest which would lead to MCP again. </p>
<p>Hope the message was clear and brief. </p>
<p>Best, </p>
<p>Umberto<br>
</p>
<p><br>
</p>
</body>
</html>