Hello,
I have a database of Stroop data that I am not sure how to analyze/ interpret. What I am looking for (I believe) are the differences in average latency between the congruent and incongruent trials. Here is a snapshot of what the variables in the database look like:
subject trialcode stimulusitem1 trialnum latency correct congruentlatency incongruentlatency
1 bluecongruent blue 3 1520 1 1520 0
1 redincongruent green 4 2425 1 1520 2425
1 blueincongruent red 5 2543 1 1520 2484
From my understanding, the congruentlatency and incongruentlatency compute average latency response times for the congruent and incongruent trials using the latency response times. For each subject, the last entries for the congruentlatency and incongruentlatency columns represent the subject's average latency times for the congruent and incongruent trials for the entire task. I've tried taking this difference for each subject, and running some analyses, but I believe this is not the appropriate method to analyze this data. Any suggestions on how to analyze this for what I'm looking for (please see above)?
Thanks in advance!
Your understanding is correct (as can easily be seen even from the bit of example data you posted).
For common analysis strategies I recommend you consult the relevant Stroop research literature. One might either compare the respective means using a t-test or an ANOVA-type analysis. For a general introduction to RT data preparation and analysis, see e.g. Lacroix & Giguere (2006).
Regards,
~Dave
Hi Dave,
Thanks for your suggestion. I have consulted some relevant Stroop research literature, and most appear to use mean response times. With that said, can you be a little more specific about:
1. what you mean by the average response times (i.e. average response time for all congruent vs incongruent trials or average difference response time for all congruent vs incongruent trials)
Thanks, again.
As far as I can see, your questions are concerned with general aspects of research methodology, data analysis and interpretation and are as such unrelated to Inquisit. Suitable analysis strategies profoundly depend on the overall design of your study (which independent variables were manipulated, was there a baseline, neutral condition, etc.) and consequently on what you can and want to know / learn from your data (given the design). There are no generally applicable answers to these questions. Thus all I can do is repeat my previous advice: Consult the relevant literature to arrive at an informed decision / strategy.
Best regards,