<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1">
</head>
<body bgcolor="#FFFFFF" text="#000000">
Dear Anahita<br>
<br>
May I suggest a different approach, which should solve your problem.<br>
<br>
You can use the LIMO EEG toolbox
(<a class="moz-txt-link-freetext" href="https://gforge.dcn.ed.ac.uk/gf/project/limo_eeg/">https://gforge.dcn.ed.ac.uk/gf/project/limo_eeg/</a>) to analyse your
data. For each participant, you build a statistical model using the
markers for faces and objects - that gives you estimates of effects
of seeing faces or objects and you can also compare within subject
faces vs objects. Once this is done, you can perform e.g. a two
samples T-test controls vs. partients on the difference (face vs
object) you computed for each subjects. The good thing here is (1)
it doesn't really matter that within subjects you don't have the
same number of (good) trials (2) it doesn't matter for the T-test if
you end up with different number of subjects (say you had to discard
2 patients because they just don't have any good trials at all) and
(3) the analysis is performed across all channels and time frames so
you don't have to do ad hoc selection of channels or time windows
(note you could also select the 'best' electrode per subject .. have
a look at our paper and tutorial)<br>
<br>
Cyril<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<table class="header-part1" border="0" cellpadding="0"
cellspacing="0" width="100%">
<tbody>
<tr>
<td>
<div class="headerdisplayname" style="display:inline;">Subject:
</div>
[Eeglablist] statistical analysis problem</td>
</tr>
<tr>
<td>
<div class="headerdisplayname" style="display:inline;">From:
</div>
anahita khorrami <a class="moz-txt-link-rfc2396E" href="mailto:asaliam20@gmail.com"><asaliam20@gmail.com></a></td>
</tr>
<tr>
<td>
<div class="headerdisplayname" style="display:inline;">Date:
</div>
15/08/2012 05:18</td>
</tr>
</tbody>
</table>
<table class="header-part2" border="0" cellpadding="0"
cellspacing="0" width="100%">
<tbody>
<tr>
<td>
<div class="headerdisplayname" style="display:inline;">To: </div>
<a class="moz-txt-link-abbreviated" href="mailto:eeglablist@sccn.ucsd.edu">eeglablist@sccn.ucsd.edu</a></td>
</tr>
</tbody>
</table>
<br>
<div class="moz-text-html" lang="x-western">Hello EEG experts<br>
I have some questions about statistical analysis. I had two groups
with 18 participants in each of them. on of the groups is typical
participants and the other one person with disorder. they examine
face comparing with other objects . finally P7 and P8 analyzed for
face. in group with disorder most of channels have value for
latency and amplitude but in other group 50% of person had zero.
now zero here is value or miss data what should I do here because
the mean of latency get less than 100 ms. I had chose the channels
which are popular for face processing and the window was 130 t0
200 ms. what should I do know. I would appreciate answer me.<br>
best regards<br>
Anahita Khorrami
</div>
<br>
<br>
<br>
<br>
<pre class="moz-signature" cols="72">--
Dr Cyril Pernet,
Academic Fellow
Brain Research Imaging Center
<a class="moz-txt-link-freetext" href="http://www.bric.ed.ac.uk/">http://www.bric.ed.ac.uk/</a>
Division of Clinical Neurosciences
University of Edinburgh
Western General Hospital
Crewe Road
Edinburgh
EH4 2XU
Scotland, UK
<a class="moz-txt-link-abbreviated" href="mailto:cyril.pernet@ed.ac.uk">cyril.pernet@ed.ac.uk</a>
tel: +44(0)1315373661
<a class="moz-txt-link-freetext" href="http://www.sbirc.ed.ac.uk/LCL/">http://www.sbirc.ed.ac.uk/LCL/</a>
<a class="moz-txt-link-freetext" href="http://www.sbirc.ed.ac.uk/cyril">http://www.sbirc.ed.ac.uk/cyril</a>
</pre>
</body>
</html>