Stroop Task Data Output


Author
Message
GradStudent5656
GradStudent5656
Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)
Group: Forum Members
Posts: 4, Visits: 27
I have run the Stroop task with keyboard inputs (English) in Inquisit 4 web.  I am now looking at the output, and want to verify the information that is presented. Below I have listed out what I believe each of the variables to be (only for the congruent trails, as I assume the information for the incongruent and controls trials is the same).

1. values.count_congruent - the number of congruent trials the participant is given
2. values.countcorrect_congruent- the number of congruent trials the participant got correct
3. expressions.congruentlatency- the average response time (in ms) of the participant on all congruent trials
4. expressions.meanRTcorr_congruent - the average response time (in ms) of the participant on the congruent trials the participant got correct
5. expressions.congruenttrials- I thought this was the percentage of the congruent trials the participant got correct.  However, the math doesn't work out.  If I divide number 2 above by number 1, that percentage doesn't equal what is presented in this cell.  I'm a bit confused.

I'd appreciate any information someone could provide on this output!  Thanks in advance for your help!

Dave
Dave
Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)
Group: Administrators
Posts: 13K, Visits: 104K
Based on your desciption, I assume you are referring to the following script:

https://www.millisecond.com/download/library/v4/Stroop/StroopWithControlKeyboard.iqx

The various variables (values, expressions, etc.) are documented inside the script:

*******************************************************************************************************************
*******************************************************************************************************************
    VALUES: automatically updated
*******************************************************************************************************************
*******************************************************************************************************************
/completed:                0 = script was not completed (script prematurely aborted); 1 = script was completed (all conditions run)
/congruencey:            1 = congruent; 2 = incongruent; 3 = control

<values>
/completed = 0
/congruency = 0
</values>

/count_congruent:                    counts all congruent trials
/countcorrect_congruent:            counts all congruent trials with a correct response
/sumrt_congruent:                    sums up the latencies of correct congruent trials
/count_incongruent:                    counts all incongruent trials
/countcorrect_incongruent:            count all incongruent trials with a correct response
/sumrt_incongruent:                    sums up the latencies of correct incongruent trials
/count_control:                        counts all control trials
/countcorrect_control:                count all control trials with a correct response
/sumrt_control:                        sums up the latencies of correct control trials


<values>
/count_congruent = 0
/countcorrect_congruent = 0
/sumrt_congruent = 0
/count_incongruent = 0
/countcorrect_incongruent = 0
/sumrt_incongruent = 0
/count_control = 0
/countcorrect_control = 0
/sumrt_control = 0
</values>

*******************************************************************************************************************
*******************************************************************************************************************
*** EXPRESSIONS*****
*******************************************************************************************************************
*******************************************************************************************************************

/ congruentlatency:                        mean latency of all congruent trials
/ incongruentlatency:                    mean latency of all incongruent trials
/ controllatency:                        mean latency of control trials

/meanRTcorr_congruent:                    mean latency of all correct congruent trials
/meanRTcorr_incongruent:                mean latency of all correct incongruent trials
/meanRtcorr_control:                    mean latency of all correct control trials

/ congruentcorrect:                        percent correct of all congruent trials
/ incongruentcorrect:                    percent correct of all incongruent trials
/ controlcorrect:                        percent correct of all control trials


<expressions>
/ congruentlatency = meanlatency(trial.bluecongruent, trial.redcongruent, trial.greencongruent, trial.blackcongruent)
/ incongruentlatency = meanlatency(trial.blueincongruent, trial.redincongruent, trial.greenincongruent, trial.blackincongruent)
/ controllatency = meanlatency(trial.bluecontrol, trial.redcontrol, trial.greencontrol, trial.blackcontrol)

/meanRTcorr_congruent = values.sumrt_congruent/values.countcorrect_congruent
/meanRTcorr_incongruent = values.sumrt_incongruent/values.countcorrect_incongruent
/meanRtcorr_control = values.sumrt_control/values.countcorrect_control

/ congruentcorrect = percentcorrect(trial.bluecongruent, trial.redcongruent, trial.greencongruent, trial.blackcongruent)
/ incongruentcorrect = percentcorrect(trial.blueincongruent, trial.redincongruent, trial.greenincongruent, trial.blackincongruent)
/ controlcorrect = percentcorrect(trial.bluecontrol, trial.redcontrol, trial.greencontrol, trial.blackcontrol)
</expressions>

I don't see the expression you mention under #5 ("expressions.congruenttrials") anywhere in the script, so I cannot speak to that particular question.

GradStudent5656
GradStudent5656
Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)Respected Member (457 reputation)
Group: Forum Members
Posts: 4, Visits: 27
Thank you for the quick response.  I assume the script you have presented is what I am using, but I went ahead and pulled the script associated with my task.  Additionally, I have downloaded a summary output file, as opposed to the raw data file.  As such, I have looked at the definitions in the script that are under "summary data" section.  The section I am referring to is highlighted that in blue in the attached script file.  The script suggests there is a countcorrect_control item, which I have highlighted in yellow.  It is supposed to count all control trials with a correct response.  However, when you look at my summary output, that label is different. I think this has something to do with the script labeling one of the columns as value.count_control instead, which is fine, but it doesn't seem to be calculating the information correct. I have highlighted that particular column label by changing the font to red in the attached script file.  Lastly, in the summary output file, that last three columns should show the percent correct for the congruent, incongruent and control trials.  I went ahead and recalculated those percentages (highlighted in green).  However, they differ from what the output shows.  Is there a reason for this difference?  Am I calculating something incorrectly?
Attachments
ScriptforStroopTask.xlsx (253 views, 19.00 KB)
SampleSummaryOutput.xlsx (258 views, 10.00 KB)
Dave
Dave
Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)Supreme Being (1M reputation)
Group: Administrators
Posts: 13K, Visits: 104K
As detailed in the comments, values.count_control reflects the overall number of control trials run

/count_control:                        counts all control trials

not the number of correct responses in control trials. There is another value that counts just the correct responses

/countcorrect_control:                count all control trials with a correct response

documented directly underneath.

Looking at the <summarydata> definition

<summarydata >
/file = "StroopwithCtrlKeyboard_summary.iqdat"
/columns = [script.startdate, script.starttime, script.subjectid, script.groupid, script.elapsedtime, values.completed,
            values.count_congruent, values.count_incongruent, values.count_control,
            values.countcorrect_congruent, values.countcorrect_incongruent, values.count_control,
            expressions.congruentlatency expressions.incongruentlatency expressions.controllatency,
            expressions.meanRTcorr_congruent, expressions.meanRTcorr_incongruent, expressions.meanRTcorr_control,
            expressions.congruentcorrect expressions.incongruentcorrect expressions.controlcorrect]
</summarydata>

it seems that the count_control value is mistakenly logged twice, while countcorrec_control is not logged. The definition was probably supposed to read

<summarydata >
/file = "StroopwithCtrlKeyboard_summary.iqdat"
/columns = [script.startdate, script.starttime, script.subjectid, script.groupid, script.elapsedtime, values.completed,
            values.count_congruent, values.count_incongruent, values.count_control,
            values.countcorrect_congruent, values.countcorrect_incongruent, values.countcorrect_control,
            expressions.congruentlatency expressions.incongruentlatency expressions.controllatency,
            expressions.meanRTcorr_congruent, expressions.meanRTcorr_incongruent, expressions.meanRTcorr_control,
            expressions.congruentcorrect expressions.incongruentcorrect expressions.controlcorrect]
</summarydata>

That unfortunately means you'll have to re-calculate the number of correct control trials from the raw data. The respective value wasn't logged to the summary data file due to the mistake in the original script outlined above.

> that last three columns should show the percent correct for the congruent, incongruent and control trials.  I went ahead and
> recalculated those percentages (highlighted in green).  However, they differ from what the output shows.  Is there a reason
> for this difference?

The slight discrepancies are probably due to various steps of rounding in the process of applying the percentcorrect() function continuously. Your calculations are fine and you can use the recorded correct and total counts to obtain percentages unencumbered by rounding.

Hope this helps.

GO

Merge Selected

Merge into selected topic...



Merge into merge target...



Merge into a specific topic ID...




Reading This Topic

Explore
Messages
Mentions
Search