Stop signal task - pre error speeding

I am working with a choice stop signal task, and need help extracting the RT before an erronous stop trial in order to investigate pre-error speeding.

This is how I extracted the goRT

for y = 1:size(eegdata.data,3)

idx = contains (eegdata.epoch(y).eventtype,'leftGoResponse') | contains(eegdata.epoch(y).eventtype,'rightGoResponse');

if (sum(idx)>0)

RT(y) = eegdata.epoch(y).eventlatency{idx};

else

RT(y)= nan;

end

end

goRT(s) = mean(RT,'omitnan')

The "names" for errounous stop trial is 'stopGoRightFA' and 'stopGoLeftFA'

So I need to extract the goRT before an errounous stop trial, and only that.

[Jeff Miller]

Is idx a single boolean just for trial y, or is it a vector? I would have thought it was just a single boolean, but sum(idx) would be unnecessary in that case, so I am confused.

[Nina Auestad]

Thanks for your comment, allow me to clear things up:

I honestly don't remember why I put the sum in there, it works without. I just put the code there for visual reasons, so that people could see what I am working with. But that's not the problem.

I need to extract the reaction time for responses that were made before unsuccessfull stop trials. What I am struggling with is that the unsuccsessful stops are in a separate epoch. So I need something that will chech if epoch 1 contains 'leftgoresponse' or 'rightgoresponse', and if epoch 2 contains 'stopGoRightFA' and 'stopGoLeftFA' , then return the RT of epoch 1 (eventlatency).

I hope that clears things up?

Answer 1

[Jeff Miller]

I think it would be cleanest to start by making a vector to indicate which trials precede unsuccessful stops, which can be done like this (if I have understood everything):

ntrials = size(eegdata.data,3);
PrecedeErrStop = false(ntrials,1);
for y=1:ntrials-1  % the last trial is not followed by anything
    PrecedeErrStop(y) = (contains(eegdata.epoch(y).eventtype,'leftGoResponse') | contains(eegdata.epoch(y).eventtype,'rightGoResponse')) ...
        & (contains(eegdata.epoch(y+1).eventtype,'stopGoRightFA') | contains(eegdata.epoch(y+1).eventtype,'stopGoLeftFA')); 
end
%PrecedeErrStop is now a vector 1:ntrials where true indicates that the
%trial does precede a stop error and false indicates that it does not.

Then you can select out the RTs that do and do not precede stop errors like this:

PrecedeErrStopRTs = eegdata.epoch(PrecedeErrStop).eventlatency{1};
NotPrecedeErrStopRTs = eegdata.epoch(~PrecedeErrStop).eventlatency{1};

The PrecedeErrStop vector will also be useful if you want to compute ERPs for the error stop trials or to exclude the error stop trials from computation of regular ERPs.

[Nina Auestad]

Thank you!!

I managed to solve it before I saw your response. But I will also try out your code. Might be easier when working with the ERPs as you say!

%% unsuccessful stops 
for y = 1:size(eegdata.data,3) - 1;
        
    idx = any(strcmp(eegdata.epoch(y).eventtype,'leftGoResponse')) | any(strcmp(eegdata.epoch(y).eventtype,'rightGoResponse')) && any(contains(eegdata.epoch(y+1).eventtype,'stopGoRightFA')) | any(contains(eegdata.epoch(y+1).eventtype,'stopGoLeftFA'));
    
    if idx;
    RT_PreError_us(y) = eegdata.epoch(y).eventlatency{2};
    else 
        RT_PreError_us(y)= nan;
    end
end
Preerror_us(s) = (mean(RT_PreError_us, 'omitnan'));