CODEWITHSUNDEEP
matlabpsychtoolbox
MGy
MGy

Reputation: 125

Using KbCheck and GetSecs to record reaction time in a conflict task in MATLAB (Psychtoolbox)

I created a Stroop-like reaction time task in MATLAB and looking at pilot results suggests there might be something wrong with my code (the congruency effect is much bigger than expected). I suspect I may be recording RTs wrong, so could anyone help me out with whether the following setup is okay?

On any given trial, two events happen (following a fixation cross): first, the target stimulus is presented for a max of 3 seconds (or until response), then the participant has to press a button to start the next trial. RT for both buttonpresses (target and trial-start button) is recorded. Here's my code:

Screen('DrawTexture', mainwin, Target);
        Screen('Flip', mainwin);
        timeStart = GetSecs;keyIsDown=0; correct=0; rt=0;
        while 1 & (GetSecs - timeStart) < 3 
                [keyIsDown, secs, keyCode] = KbCheck;
                FlushEvents('keyDown');
                if keyIsDown
                    nKeys = sum(keyCode);
                    if nKeys==1
                        if keyCode(Left)||keyCode(Right)||keyCode(Down)||keyCode(Up)
                            rt = 1000.*(GetSecs-timeStart);
                            keypressed=find(keyCode);
                            Screen('Flip', mainwin);
                            if ... [I removed some irrelevant ERROR feedback related code here]...
                        elseif keyCode(escKey)
                            ShowCursor; fclose(outfile);  Screen('CloseAll'); return
                        end
                        keyIsDown=0; keyCode=0;
                    end
                else
                    keypressed = 0; %the resp column in output is 0 if no button is pressed
                end
        end
        if keypressed == 0 %indicates timeout errors
            DrawFormattedText(mainwin, 'TOO SLOW', 'center', 'center', errorcol);
            Screen('Flip', mainwin);
            WaitSecs(1);
        end
        Screen('DrawTexture', mainwin, press5);
        Screen('Flip', mainwin);
        keyIsDown=0; timeSt = GetSecs;
        while 1
            [keyIsDown, secs, keyCode] = KbCheck;
            if keyIsDown
                if keyCode(MoveOn)
                    pause_rt = 1000.*(secs - timeSt);
                    break ;
                elseif keyCode(escKey)
                    ShowCursor;
                    fclose(outfile);
                    Screen('CloseAll');
                    return;
                end
            end
        end

My questions: the command GetSecs gets the time whenever it's called, right? So GetSecs - timeStart is an okay way of calculating RT - but so is secs - timeSt (as seen for the second stimulus), as secs is the time KbCheck returns for the buttonpress. The two methods are largely equivalent (with GetSecs - timeStart maybe slightly overestimating RT), is that correct?

My worry here is that RT estimates for the target on the NEXT trial might be influenced by the second button press RT of the PREVIOUS trial. Do you see any evidence of that?

Upvotes: 0

Views: 1218

Answers (1)

DMR
DMR

Reputation: 1473

You are correct that GetSecs returns the time whenever it is called. But, although not the cause of your error, calling GetSecs after each window Flip isn't necessary, because the Flip function returns the estimate of stimulus onset, as the second output of the function. So for example, instead of:

Screen('Flip', mainwin); timeStart = GetSecs;

You can just use

[~, timeStart] = Screen('Flip', mainwin);

I don't notice anything obviously wrong in the code, when you say the RTs seem incorrect, do they seem too fast, or too slow? Personally I would use KbReleaseWait after collecting RT to wait until the Key has been released, but perhaps FlushEvents is doing a similar thing here.

Upvotes: 1

Related Questions