[FieldTrip] Information about MANOVA function

[Davide Simeone]

Thank you so much! You understood correctly, my doubt had arisen due to the name of the function, but now I understand.

Davide

From: fieldtrip <fieldtrip-bounces at science.ru.nl> on behalf of Schoffelen, J.M. (Jan Mathijs) via fieldtrip <fieldtrip at science.ru.nl>
Date: Wednesday, 5 June 2024 at 10:30 AM
To: FieldTrip discussion list <fieldtrip at science.ru.nl>
Cc: Schoffelen, J.M. (Jan Mathijs) <janmathijs.schoffelen at donders.ru.nl>
Subject: Re: [FieldTrip] Information about MANOVA function
Hi Davide,

I may have misunderstood the question behind your question, but indeed FieldTrip’s statfuns treat each data point (channel-time-frequency) as a single ‘univariate' variable, and compute a mass univariate statistic.

So, where you say: '-why an F-score is obtained for each dependent variable;’ -> this is just the way that the function works, per definition
and where you say: '-why the other dependent variables are not included in this computation’ -> well, they are, but they are indeed all treated as univariate quantities.

I think that the ‘multivariate’ in the name of the function refers to the multivariate nature of the design, and not to the intention of combining the multivariate data into a single test statistic.

Perhaps Eric Maris can chime in here, he wrote the original function, I never use it myself.

Best wishes,
Jan-Mathijs



On 27 May 2024, at 19:48, Davide Simeone via fieldtrip <fieldtrip at science.ru.nl<mailto:fieldtrip at science.ru.nl>> wrote:

Dear community,
I would like to ask a theoretical question. In the ‘ft_statfun_depsamplesFmultivariate’ function, the statistic computation is made by a for cycle on each sample, that would correspond, in a classical problem following the tutorial ‘Cluster-based permutation tests on time-frequency data’, to a ‘time frequency channel’ value, in the most complex case. In a within subject experiment, I would think about each of these values as dependent variables (but maybe I’m already wrong here). Assuming that my previous affirmation is correct, I don’t understand
-why an F-score is obtained for each dependent variable;
-why the other dependent variables are not included in this computation.
In simpler words, it seems that the function acts as doing multiple ANOVA tests, instead of a multivariate one. My comprehension of these statistics tests is limited, so I could be wrong about the assumptions that I have done, but I would like to improve my understanding and be clarified even more if that is the case.
Greetings,
Davide


_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20240605/b0952903/attachment.htm>