[Eeglabnews] EEGLAB 9.0 now available

[Arnaud Delorme]

EEGLAB version 9.0.0.1b is now available from http://www.sccn.ucsd.edu/eeglab/ 

This update is strongly recommended as it fixes some important issues:

1. Because of an error to speed up the time-frequency decomposition functions, FFT frequency have been slightly inaccurate from the period of June 2009 to this present release. This problem only occurred when selecting custom window size (not using the default window size which is a power of 2). Frequencies where overestimated by a factor of 0% to 50% (the exact number depending on the size of the window). You need not be concerned by this bug unless you selected manually a window size which was not a power of 2 (thanks to Wu Quian).
 
2. A problem remains with the bootstrap function for paired data. The permutation function should be used instead. This problems arises because of resampling method we used. It leads to overestimation of significance by a factor of about 2 compared to permutation. We have not found a solution yet for this problem but have placed a warning when you choose to use bootstrap in the graphic interface. Bootstrap for unpaired data is functional and recommended (thanks to Erik Muelle).

In addition, a major upgrade has been introduced since EEGLAB v8 -- a new STUDY.design structure and facilities.

1. Studies may now have customized independent variable names.

2. More importantly, multiple statistical designs may be used to analyze a single study, each design involving either all or only some of the datasets associated with the study.

3. The STUDY.design facility also removes the constraint of needing to build a separate dataset for each experimental condition.

For more information see http://sccn.ucsd.edu/wiki/Chapter_03:_Working_with_STUDY_designs

The list of all the version updates is available on the EEGLAB wiki.

Visit  http://sccn.ucsd.edu/wiki/EEGLAB_revision_history

The EEGLAB team

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://sccn.ucsd.edu/pipermail/eeglabnews/attachments/20100903/b8467d10/attachment.html