[Eeglablist] ANOVA and Bonferroni_eeglab vs. SPSS

[Makoto Miyakoshi]

Dear Johanna,

Thanks for clarification.

> Yes, I've used a 4x2 mixed desgin. I tested 10 subjects and each of them
passed randomised through each condition. My 4 independent variables were:

I'm not still sure if the 4x2 design is all within-subject design (i.e.,
repeated measures) or 'mixed' as you say (i.e., between-subject condition
for mental vs. actual dancing). The design looks like within-subject to me,
but you say 'mixed'...

Assuming that all conditions are within-subject design, when you use
statcond(), use 'paired', {'on' 'on'}, 'method', 'param' to perform the
standard 2-way ANOVA with repeated measures. The result should match that
of SPSS.

Makoto

On Mon, Jul 23, 2018 at 1:26 AM Wind, Johanna <jwind01 at uni-mainz.de> wrote:

> Dear Makoto,
>
>
> thanks for your fast answering!
>
>
> >However, as far as I know, EEGLAB's ANOVA may have limitation in mixed
> design i.e., mixture of repeated and non-repeated measures.<
>
>
> Okay, Maybe that's the "problem", cause I computed a repeated ANOVA.
>
>
>
>
> >If your 4x2 factorial deisgn mixed design? Also, what did you test? Mean
> amplitude of a window, or what? Give me more info and I may be able to help
> you.<
>
>
> Yes, I've used a 4x2 mixed desgin. I tested 10 subjects and each of them
> passed randomised through each condition. My 4 independent variables were:
>
>    1. 15 min. dancing with music
>    2. 15 min. dancing without music
>    3. 15 min. sitting and dance mentally with music
>    4. 15 min. sitting and dance mentally with music
>
> Before and after each condition were the EEG measurements (resting
> condition; 2 min. eyes-open, 2 min. eyes-closed). I tested the mean
> amplitude of the resting conditions.
>
> Hope this was explicitly enough?
>
> Thanks a lot in advance!
>
> Johanna
>
>
>
>
> Johanna Wind
>
> wissenschaftliche Hilfskraft
>
> Johannes Gutenberg-Universität Mainz
>
> Institut für Sportwissenschaft
>
> Abt. Sportpädagogik/ -Ethik
>
> Raum: 00-127
>
> Tel.: 06131-39-28254
>
> E-Mail: jwind01 at uni-mainz.de
>
>
> ------------------------------
> *Von:* Makoto Miyakoshi <mmiyakoshi at ucsd.edu>
> *Gesendet:* Freitag, 6. Juli 2018 21:11:35
> *An:* Wind, Johanna
> *Cc:* EEGLAB List
> *Betreff:* Re: [Eeglablist] ANOVA and Bonferroni_eeglab vs. SPSS
>
> Dear Johanna,
>
> > Does eeglab use the same computation of ANOVA and Bonferroni-correction
> like SPSS does?
>
> EEGLAB uses statcond() for parametric (t-test and ANOVA) and
> non-parametric (permutation, bootstrap). For multiple comparison
> correction, it has fdr() and bonf_holm(). I don't use SPSS but the
> algorithm should be the same. However, as far as I know, EEGLAB's ANOVA may
> have limitation in mixed design i.e., mixture of repeated and non-repeated
> measures.
>
> > For clearer understanding here a short overview of my study design:
> There were 4 conditions and 2 testing times (pretest and posttest) with 19
> electrodes. I computed a pairwise comparison between the conditions and the
> testing times. There were significant differences between the conditions in
> SPSS, but not so in eeglab.
>
> Oh boy, this is not cleaner to me unfortunately!
> Did you perform ANOVA, or what? If your 4x2 factorial deisgn mixed design?
> Also, what did you test? Mean amplitude of a window, or what? Give me more
> info and I may be able to help you.
>
> Makoto
>
>
> On Fri, Jul 6, 2018 at 10:33 AM Wind, Johanna <jwind01 at uni-mainz.de>
> wrote:
>
>>
>>
>> Hello everybody,
>>
>>
>>
>> I‘ve got a question in relation to the eeglab output. Does eeglab use the
>> same computation of ANOVA and Bonferroni-correction like SPSS does? When I
>> transfer the specdata from eeglab to SPSS and compute an ANOVA with post
>> hoc tests in order to get pairwise comparisons between my conditions, there
>> were given significant differences. But in the pairwise comparison of the
>> conditions in eeglab are not all electrodes significant like shown in SPSS.
>>
>>
>>
>> For clearer understanding here a short overview of my study design:
>>
>> There were 4 conditions and 2 testing times (pretest and posttest) with
>> 19 electrodes. I computed a pairwise comparison between the conditions and
>> the testing times. There were significant differences between the
>> conditions in SPSS, but not so in eeglab.
>>
>>
>>
>>
>>
>>
>>
>> It would be awesome, if you can reply my questions.
>>
>>
>>
>>
>>
>> Best regards,
>>
>>
>>
>> Johanna Wind
>>
>> wissenschaftliche Mitarbeiterin
>>
>> Johannes Gutenberg-Universität Mainz
>>
>> Institut für Sportwissenschaft
>>
>> Abt. Trainings- und Bewegungswissenschaften
>>
>> Raum: 00-113
>>
>> Tel.: 06131-24560
>>
>> Albert Schweitzer-Straße 22
>>
>> E-Mail: jwind01 at uni-mainz.de
>>
>> Sprechstunde nach Vereinbarung
>>
>>
>> _______________________________________________
>> 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
>


-- 
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/20180731/f49dbdab/attachment.html>