By shezza098 - 8/15/2018
Hello!
I am very new to Stroop task and have only just used it in my thesis (an emotional Stroop task) so I apologise if this question has an obvious answer.
I have collected my data and I was wondering how to go about analysing it.
From what I understand I will have to calculate an interference score (e.g., interference score = neutral word - negative word). For the neutral word score, do I just average the reaction times for all the neutral words and use that to calculate the interference score for each negative word and then average that to get a total score for negative word for each participant? Also how do the correct and incorrect responses fit into this?
I am sorry if this is obvious! I am new to Stroop.
Thanks!
Sherry
|
By Dave - 8/15/2018
+xHello! I am very new to Stroop task and have only just used it in my thesis (an emotional Stroop task) so I apologise if this question has an obvious answer. I have collected my data and I was wondering how to go about analysing it. From what I understand I will have to calculate an interference score (e.g., interference score = neutral word - negative word). For the neutral word score, do I just average the reaction times for all the neutral words and use that to calculate the interference score for each negative word and then average that to get a total score for negative word for each participant? Also how do the correct and incorrect responses fit into this? I am sorry if this is obvious! I am new to Stroop. Thanks! Sherry
Assuming you used this script, https://www.millisecond.com/download/library/v5/stroop/emotionalstroop/emotionalstroopwithkeyboardinput.manual , bias scores are already calculated and provided in your summary data file. See the expressions.Bias_Agg to expressions.Bias_Pos variables in the output.
They are calculated as the difference in mean latency between the emotional category minus the neutral category, as you can see here:
<expressions>
/percCorrect_Agg = (values.correct_aggressionwords/trial.aggression.trialcount) * 100
/percCorrect_Neg = (values.correct_negativewords/trial.negativeword.trialcount) * 100
/percCorrect_Neu = (values.correct_neutralwords/trial.neutralword.trialcount) * 100
/percCorrect_Col = (values.correct_colorwords/trial.colorword.trialcount) * 100
/percCorrect_Pos = (values.correct_positivewords/trial.positiveword.trialcount) * 100
/meanRT_Agg = values.sumrt_agg/values.correct_aggressionwords
/meanRT_Neg = values.sumrt_neg/values.correct_negativewords
/meanRT_Neu = values.sumrt_neu/values.correct_neutralwords
/meanRT_Pos = values.sumrt_pos/values.correct_positivewords
/meanRT_Col = values.sumrt_col/values.correct_colorwords
/Bias_Agg = expressions.meanrt_agg - expressions.meanrt_neu
/Bias_Neg = expressions.meanrt_neg - expressions.meanrt_neu
/Bias_Pos = expressions.meanrt_pos - expressions.meanrt_neu
/Bias_Col = expressions.meanrt_col - expressions.meanrt_neu
</expressions>
As you can also see in the user manual as well as in the above, only latencies stemming from *correct* responses are considered.
|
By shezza098 - 8/15/2018
Hi Dave,
Thanks so much for your help!! I realised my mistake, I modified the Stroop file to include my categories, but unfortunately didn’t change the summary file section in the instructions so it still reflects the original Stroop task (rookie mistake). I suppose I will just have to calculate the results from the raw data file.
Thanks!
Sherry
|
By shezza098 - 8/15/2018
Hi Dave,
Just to confirm, the calculation for the raw data would be
MeanRT= sum of latency scores for aggression words/ # of correct responses
Bias for aggressive words = meanRT for agg words - mean RT for neutral words
Is that correct?
Thanks again! I appreciate your help!
Sherry
|
By Dave - 8/16/2018
+xHi Dave, Just to confirm, the calculation for the raw data would be MeanRT= sum of latency scores for aggression words/ # of correct responses Bias for aggressive words = meanRT for agg words - mean RT for neutral wordsIs that correct? Thanks again! I appreciate your help! Sherry
Yes, that is correct, but you only sum up the latencies for *correct* trials, you toss out all incorrect ones.
|
By shezza098 - 8/16/2018
Yeah great!! Thanks again :)
|
|