[FieldTrip] Fwd: correlation with behavioural data and cluster statistics

Tue Apr 2 11:57:36 CEST 2019

Dear Arjen,

 if I use matlab function "corr" I could get significant correction between behavioral data and EEG data calculated with predefined time, frequency and channels.

However, no significant cluter, while using either correlationT or ft_statfun_indepsamplesregrT even averaged within the same time, frequency and channels.

I could not understand it. Could you guide me?

Best,

Juan

________________________________
From: fieldtrip <fieldtrip-bounces at science.ru.nl> on behalf of Arjen Stolk <a.stolk8 at gmail.com>
Sent: Sunday, March 17, 2019 7:12 AM
To: FieldTrip discussion list
Subject: Re: [FieldTrip] Fwd: correlation with behavioural data and cluster statistics

Dear Juan,

Your code looks fine to me. Indeed, also with ft_statfun_indepsamplesregrT, the behavioral vector goes into the design matrix.

All p-values computed with the montecarlo method correspond to one tail and, thus, you would need to correct for this one-sided test approach if you intended to do a two-sided test. Either cfg.correcttail = 'alpha' or 'prob' will achieve this, as further explained here<https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.fieldtriptoolbox.org%2Ffaq%2Fwhy_should_i_use_the_cfg.correcttail_option_when_using_statistics_montecarlo%2F&data=02%7C01%7C%7C6482c0ed5b854521095608d6aaa1afee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636884008747439566&sdata=cCUJE92o9BrAd35xx36cP0fkXukAUhsNW%2BEt4xMnwCM%3D&reserved=0>.

Have a great weekend,
Arjen

On Tue, Mar 12, 2019 at 7:35 AM Juan Lei <juan.lei at hotmail.com<mailto:juan.lei at hotmail.com>> wrote:
Dear Arjen,

if I use ft_statfun_indepsamplesregrT, are all the Parameters the same as correlationT method ? Below are all my parameters used. Are they correct? Another question is whether I need the "cfg.correcttail  =  'alpha'" for the indepsamplesregrT methoed?

   cfg=[];
   cfg.channel          = {'*'};
   cfg.frequency        = [15 35];
   cfg.avgoverfreq      ='no';
   cfg.latency          = [-0.7 0];
   cfg.avgoverchan      = 'no';
   cfg.avgovertime      = 'no';
   cfg.parameter        = 'powspctrm';
   cfg.method           = 'montecarlo';
   cfg.alpha            = 0.05;
   cfg.numrandomization = 1000;
   cfg.correctm         = 'cluster';
   cfg.clusteralpha     = 0.05;
   cfg.clusterstatistic = 'maxsum';
   cfg.correcttail      =  'alpha';
   cfg.tail             = 0;
   cfg.clustertail      = 0;
   cfg.minnbchan        = 2;
   cfg.neighbours       = neighbours;
   cfg.statistic        ='ft_statfun_indepsamplesregrT';
   cfg.ivar             = 1;
   cfg.design = [];
   cfg.design(1,:) = xa;% xa is my behavioral vector. Each subject has one behavioral value.
   freq_stat = ft_freqstatistics(cfg,GA_RI);% GA_RI is my time frequency data averaged across group but keeping individuals.

Best regards,

Juan Lei

________________________________
From: fieldtrip <fieldtrip-bounces at science.ru.nl<mailto:fieldtrip-bounces at science.ru.nl>> on behalf of Valeriya Belyaeva <vabelyaeva94 at gmail.com<mailto:vabelyaeva94 at gmail.com>>
Sent: Tuesday, February 19, 2019 7:11 PM
To: FieldTrip discussion list
Subject: Re: [FieldTrip] Fwd: correlation with behavioural data and cluster statistics

Thank you a lot, Arjen.

Best regards,
Valeriya

вт, 19 февр. 2019 г. в 18:44, Arjen Stolk <a.stolk8 at gmail.com<mailto:a.stolk8 at gmail.com>>:
Hi Valeriya,

That looks to be correct use of the correlationT method. The behavioral data goes into the design matrix, as you did.

Best,
Arjen

On Feb 19, 2019, at 6:35 AM, Valeriya Belyaeva <vabelyaeva94 at gmail.com<mailto:vabelyaeva94 at gmail.com>> wrote:

---------- Forwarded message ---------
From: Valeriya Belyaeva <vabelyaeva94 at gmail.com<mailto:vabelyaeva94 at gmail.com>>
Date: ср, 13 февр. 2019 г. в 16:28
Subject: correlation with behavioural data and cluster statistics
To: <fieldtrip at science.ru.nl<mailto:fieldtrip at science.ru.nl>>

Dear fieldtrip developers,

I have a question about correlation of neuronal and behavioural data. You have discussed this question here: http://www.fieldtriptoolbox.org/faq/how_can_i_test_for_correlations_between_neuronal_data_and_quantitative_stimulus_and_behavioural_variables/<https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.fieldtriptoolbox.org%2Ffaq%2Fhow_can_i_test_for_correlations_between_neuronal_data_and_quantitative_stimulus_and_behavioural_variables%2F&data=02%7C01%7C%7C6482c0ed5b854521095608d6aaa1afee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636884008747449571&sdata=qic0njE42iEFnLzLpjI5pOH3Qixs6BfGYLlsS2yuVPk%3D&reserved=0>

I have taken the code, which was given as an example and added cluster-based correction method. My goal is to perform correlation between neuronal and behavioural data and then to account for multiple comparisons with cluster statistics.

Could you, please, tell me, whether this script does it?

Here is the script:

% example script

cfg = [];

cfg.statistic        = 'ft_statfun_correlationT';

cfg.method           = 'montecarlo';

cfg.numrandomization = 1000;

cfg.correctm         = 'cluster';

cfg.clusteralpha     = 0.025;

cfg_neighb.method = 'triangulation';

cfg_neighb.layout = 'acticap_62';

cfg.neighbours = ft_prepare_neighbours(cfg_neighb, data_brain{:});

n1 = 42;    % n1 is the number of subjects

design(1,1:n1)       = behav_data(:,1); % mean reaction time by subject

cfg.design           = design;

cfg.ivar             = 1;

stat = ft_freqstatistics(cfg, data_brain{:});

Thank you a lot in advance!
--
Best regards,
Valeriya Belyaeva
HSE, Cognitive science and technologies

--
Best regards,
Valeriya Belyaeva
HSE, Cognitive science and technologies

_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmailman.science.ru.nl%2Fmailman%2Flistinfo%2Ffieldtrip&data=02%7C01%7C%7C6482c0ed5b854521095608d6aaa1afee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636884008747459581&sdata=N8cEFn91nSkR8l0OA%2B6%2BqQYs%2Bcnx9ZJytZMu8uQFDYo%3D&reserved=0>
https://doi.org/10.1371/journal.pcbi.1002202<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdoi.org%2F10.1371%2Fjournal.pcbi.1002202&data=02%7C01%7C%7C6482c0ed5b854521095608d6aaa1afee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636884008747469586&sdata=dr%2FO6kt9rbRIPu3GpQk1D9%2BVuxMfg5OxPJzNQgoZheQ%3D&reserved=0>
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmailman.science.ru.nl%2Fmailman%2Flistinfo%2Ffieldtrip&data=02%7C01%7C%7C6482c0ed5b854521095608d6aaa1afee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636884008747489602&sdata=q91RtnrucbD%2Fd%2Fn9v1f9EyJihjXCUSbstmU9gYK66Wo%3D&reserved=0>
https://doi.org/10.1371/journal.pcbi.1002202<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdoi.org%2F10.1371%2Fjournal.pcbi.1002202&data=02%7C01%7C%7C6482c0ed5b854521095608d6aaa1afee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636884008747499607&sdata=q3op8RD6C6E3WFxcQ2I8%2Fg7bkfzX6Vyc5jsGHfWOfPk%3D&reserved=0>

--
Best regards,
Valeriya Belyaeva
HSE, Cognitive science and technologies

_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmailman.science.ru.nl%2Fmailman%2Flistinfo%2Ffieldtrip&data=02%7C01%7C%7C6482c0ed5b854521095608d6aaa1afee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636884008747510970&sdata=wHjEE%2B8%2FqnJYLVLjx7OmC9bBLhhV84h4ZttZtC5pUdo%3D&reserved=0>
https://doi.org/10.1371/journal.pcbi.1002202<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdoi.org%2F10.1371%2Fjournal.pcbi.1002202&data=02%7C01%7C%7C6482c0ed5b854521095608d6aaa1afee%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636884008747519617&sdata=gVJd%2BEAz4hsZKd6v2lDgXG98gsnhELRu%2B9%2BHU80jAFc%3D&reserved=0>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20190402/8f6963ad/attachment-0001.html>
-------------- next part --------------
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202