I am new to PsPM and I'm using it to analyse SCR with a DCM. With peak scoring, SCR is usually square root transformed. I was wondering if this is also done in PsPM by default? If yes, is this done before downsampling, filtering and normalizing, or after? Thank you in advance!
Tanja
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
SCR amplitude or magnitude measures from peak-scoring analysis are sometimes square-root or log-transformed to render their distribution more "normal", similar to a link function in a generalised linear model. However, I am not aware of any systematic investigation under which conditions this is meaningful or necessary - and the choice often seems to be based on tradition. This is why PsPM does not transform the estimates and leaves this to the user.
In peak-scoring (and DCM), the distribution of SCR amplitude estimates is skewed because they are constrained to be positive. However, this by itself is not a problem if you are testing condition differences, which can be negative or positive.
I would suggest to do what you would do for any DV in any statistical model: run the model and test for heteroscedasticity of the residuals.
Hope this helps
Dominik
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi everyone,
I am new to PsPM and I'm using it to analyse SCR with a DCM. With peak scoring, SCR is usually square root transformed. I was wondering if this is also done in PsPM by default? If yes, is this done before downsampling, filtering and normalizing, or after? Thank you in advance!
Tanja
Hi Tanja
SCR amplitude or magnitude measures from peak-scoring analysis are sometimes square-root or log-transformed to render their distribution more "normal", similar to a link function in a generalised linear model. However, I am not aware of any systematic investigation under which conditions this is meaningful or necessary - and the choice often seems to be based on tradition. This is why PsPM does not transform the estimates and leaves this to the user.
In peak-scoring (and DCM), the distribution of SCR amplitude estimates is skewed because they are constrained to be positive. However, this by itself is not a problem if you are testing condition differences, which can be negative or positive.
I would suggest to do what you would do for any DV in any statistical model: run the model and test for heteroscedasticity of the residuals.
Hope this helps
Dominik
Thank you so much for your answer, this clarified everything for me!