Data analysis for Stroop task with keyboard


Data analysis for Stroop task with keyboard

Author
Message
ajilime
ajilime
New Member
New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)

Group: Forum Members
Posts: 3, Visits: 11
The only values I have that can be used are expression.meanRT, expression.propcorrect, expression.meanRTcorr_incongruent, and expression.propcorrect_incongruent. Does anyone have any idea how to sum up this data in order to receive one score? Any papers that might have used similar method would be very much appreciated. 

Thanks
Dave
Dave
Supreme Being
Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)

Group: Administrators
Posts: 9.5K, Visits: 43K
ajilime - Tuesday, May 7, 2019
The only values I have that can be used are expression.meanRT, expression.propcorrect, expression.meanRTcorr_incongruent, and expression.propcorrect_incongruent. Does anyone have any idea how to sum up this data in order to receive one score? Any papers that might have used similar method would be very much appreciated. 

Thanks

You'll typically look for a significant difference between the reaction time for congruent trials (expressions.meanRTcorr_congruent) and incongruent trials (expressions.meanRTcorr_incongruent); the expectation is that participants perform faster in congruent trials than in incongruent trials. That is, your independent variable is congruency (color word printed in matching color vs color word printed in non-matching color) and your dependent variable is the response time. You can do the same with correct proportions as your dependent variable, where the expectation is that participants peform better (i.e. fewer errors) in congruent trials than in incongruent trials (more errors).

ajilime
ajilime
New Member
New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)

Group: Forum Members
Posts: 3, Visits: 11
Dave - Tuesday, May 7, 2019
ajilime - Tuesday, May 7, 2019
The only values I have that can be used are expression.meanRT, expression.propcorrect, expression.meanRTcorr_incongruent, and expression.propcorrect_incongruent. Does anyone have any idea how to sum up this data in order to receive one score? Any papers that might have used similar method would be very much appreciated. 

Thanks

You'll typically look for a significant difference between the reaction time for congruent trials (expressions.meanRTcorr_congruent) and incongruent trials (expressions.meanRTcorr_incongruent); the expectation is that participants perform faster in congruent trials than in incongruent trials. That is, your independent variable is congruency (color word printed in matching color vs color word printed in non-matching color) and your dependent variable is the response time. You can do the same with correct proportions as your dependent variable, where the expectation is that participants peform better (i.e. fewer errors) in congruent trials than in incongruent trials (more errors).

Thank you for your reply. The problem is, with the data that I have I can only use reaction time for all trials (congruent, incongruent, and control) and reaction time for incongruent trials. Is there any way to look for a significant there?
Dave
Dave
Supreme Being
Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)

Group: Administrators
Posts: 9.5K, Visits: 43K
ajilime - Tuesday, May 7, 2019
Dave - Tuesday, May 7, 2019
ajilime - Tuesday, May 7, 2019
The only values I have that can be used are expression.meanRT, expression.propcorrect, expression.meanRTcorr_incongruent, and expression.propcorrect_incongruent. Does anyone have any idea how to sum up this data in order to receive one score? Any papers that might have used similar method would be very much appreciated. 

Thanks

You'll typically look for a significant difference between the reaction time for congruent trials (expressions.meanRTcorr_congruent) and incongruent trials (expressions.meanRTcorr_incongruent); the expectation is that participants perform faster in congruent trials than in incongruent trials. That is, your independent variable is congruency (color word printed in matching color vs color word printed in non-matching color) and your dependent variable is the response time. You can do the same with correct proportions as your dependent variable, where the expectation is that participants peform better (i.e. fewer errors) in congruent trials than in incongruent trials (more errors).

Thank you for your reply. The problem is, with the data that I have I can only use reaction time for all trials (congruent, incongruent, and control) and reaction time for incongruent trials. Is there any way to look for a significant there?

> The problem is, with the data that I have I can only use reaction time for all trials (congruent, incongruent, and control) and reaction time for incongruent trials.

Why? Even if something went wrong with your summary data, I assume you have the raw data file and you can thus calculate separate means for congruent, incongruent and control trials.

ajilime
ajilime
New Member
New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)New Member (22 reputation)

Group: Forum Members
Posts: 3, Visits: 11
Dave - Tuesday, May 7, 2019
ajilime - Tuesday, May 7, 2019
Dave - Tuesday, May 7, 2019
ajilime - Tuesday, May 7, 2019
The only values I have that can be used are expression.meanRT, expression.propcorrect, expression.meanRTcorr_incongruent, and expression.propcorrect_incongruent. Does anyone have any idea how to sum up this data in order to receive one score? Any papers that might have used similar method would be very much appreciated. 

Thanks

You'll typically look for a significant difference between the reaction time for congruent trials (expressions.meanRTcorr_congruent) and incongruent trials (expressions.meanRTcorr_incongruent); the expectation is that participants perform faster in congruent trials than in incongruent trials. That is, your independent variable is congruency (color word printed in matching color vs color word printed in non-matching color) and your dependent variable is the response time. You can do the same with correct proportions as your dependent variable, where the expectation is that participants peform better (i.e. fewer errors) in congruent trials than in incongruent trials (more errors).

Thank you for your reply. The problem is, with the data that I have I can only use reaction time for all trials (congruent, incongruent, and control) and reaction time for incongruent trials. Is there any way to look for a significant there?

> The problem is, with the data that I have I can only use reaction time for all trials (congruent, incongruent, and control) and reaction time for incongruent trials.

Why? Even if something went wrong with your summary data, I assume you have the raw data file and you can thus calculate separate means for congruent, incongruent and control trials.

I took over the project from another researcher and I am not sure if I would be able to access/find all of their data. 
Dave
Dave
Supreme Being
Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)Supreme Being (610K reputation)

Group: Administrators
Posts: 9.5K, Visits: 43K
ajilime - Wednesday, May 8, 2019
Dave - Tuesday, May 7, 2019
ajilime - Tuesday, May 7, 2019
Dave - Tuesday, May 7, 2019
ajilime - Tuesday, May 7, 2019
The only values I have that can be used are expression.meanRT, expression.propcorrect, expression.meanRTcorr_incongruent, and expression.propcorrect_incongruent. Does anyone have any idea how to sum up this data in order to receive one score? Any papers that might have used similar method would be very much appreciated. 

Thanks

You'll typically look for a significant difference between the reaction time for congruent trials (expressions.meanRTcorr_congruent) and incongruent trials (expressions.meanRTcorr_incongruent); the expectation is that participants perform faster in congruent trials than in incongruent trials. That is, your independent variable is congruency (color word printed in matching color vs color word printed in non-matching color) and your dependent variable is the response time. You can do the same with correct proportions as your dependent variable, where the expectation is that participants peform better (i.e. fewer errors) in congruent trials than in incongruent trials (more errors).

Thank you for your reply. The problem is, with the data that I have I can only use reaction time for all trials (congruent, incongruent, and control) and reaction time for incongruent trials. Is there any way to look for a significant there?

> The problem is, with the data that I have I can only use reaction time for all trials (congruent, incongruent, and control) and reaction time for incongruent trials.

Why? Even if something went wrong with your summary data, I assume you have the raw data file and you can thus calculate separate means for congruent, incongruent and control trials.

I took over the project from another researcher and I am not sure if I would be able to access/find all of their data. 

I see. I don't think a meaningful analysis is possible with the limited data / variables you have, so trying to find / get access to the remaining data would be very much worthwhile. If that data was collected with Inquisit Web and may still be available in some account you don't have access to, we can possibly help locate it. I would, however, need information like account name, names and/or email addresses of the researcher(s) who owned and/or had access to the account -- basically anything you might know -- for that. If you have any such information, please feel free to send me a private forum message.

GO


Reading This Topic


Login
Existing Account
Email Address:


Password:


Social Logins

Select a Forum....






Millisecond Forums


Search