I am wondering about my CDA analysis export. As far as I understood I just need to take the average CDA.SCR per condition (of my events) and that is what I would work with in my statistics. But I just realized that sometimes there is a CDA.SCR value, but there was actually no response found (CDA.nSCR is zero) for a certain event. How do I deal with that? Do I still include that value for my exported mean? Or should I exclude this when I calculate it?
Another related question is: How do I deal with subjects, that do not show frequent responses, is there any rule of thumb or cutoff, how many identified SCR one needs?
thank you,
Katharina
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
CDA.nSCR uses a minimum amplitude criterion for what counts as an SCR. So SCR with lower amplitude will still contribute phasic activation. As this criterion is more or less arbitrary and CDA.SCR considers the full time course rather than discretionary responses, I don't think this is much a problem.
Different criteria for excluding non-responder have been proposed often considering a minimum number of responses per condition (e.g. 50%, but other studies huse other thresholds).
Mathias
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi to all,
I am wondering about my CDA analysis export. As far as I understood I just need to take the average CDA.SCR per condition (of my events) and that is what I would work with in my statistics. But I just realized that sometimes there is a CDA.SCR value, but there was actually no response found (CDA.nSCR is zero) for a certain event. How do I deal with that? Do I still include that value for my exported mean? Or should I exclude this when I calculate it?
Another related question is: How do I deal with subjects, that do not show frequent responses, is there any rule of thumb or cutoff, how many identified SCR one needs?
thank you,
Katharina
CDA.nSCR uses a minimum amplitude criterion for what counts as an SCR. So SCR with lower amplitude will still contribute phasic activation. As this criterion is more or less arbitrary and CDA.SCR considers the full time course rather than discretionary responses, I don't think this is much a problem.
Different criteria for excluding non-responder have been proposed often considering a minimum number of responses per condition (e.g. 50%, but other studies huse other thresholds).
Mathias