[Eeglablist] Including partial data

[Makoto Miyakoshi]

Dear Dave,

How to deal with missing data seems a simple but profound issue, and there
are range of solutions from the simplest to the most sophisticated. If you
know how to deal with missing data properly using these methods, that's the
best. Otherwise, it's probably better to give up the two subjects.

Makoto

On Fri, Jul 15, 2016 at 11:32 AM, Jenson, Dave <djenson1 at uthsc.edu> wrote:

> All,
>
>
> I have a 4x2 design (4 conditions, 2 groups), and am only interested in
> between-group differences.  I have a few subjects (2 out of 25) from a
> difficult to recruit population who, due to failure to follow instructions,
> only provided data in 3 of the 4 conditions.  In the STUDY module I am
> using permutation stats with a cluster-based correction to determine group
> differences.
>
>
> I want to know if it is valid to include the data from these partial
> contributors, or if that will lead to unreliable results.  I'm under the
> impression that since permutation stats use random sampling to generate a
> surrogate distribution, missing data shouldn't be a problem.  However, if
> EEGLAB fills in zeros or NaNs for missing data, that could potentially
> compromise the results.  Has anyone tackled this problem before?
>
>
> -Dave
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu
>



-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20160725/467f0c70/attachment.html>