Our data is from two scalp EEG channels (C3 and C4), which correspond to the sensorimotor cortex on each side of the brain. We are looking at how effect size of a waveform metric differs between patients on and off medication (n=15). We also want to look at a subset of the patients (n=9 for C3 and n=8 for C4) and compare it to the effect size of all the patients. We'd like to pool the data i.e. not have to report effect size for C3 and C4 separately, if possible. Questions: What effect size would be best to use (Cohen's D, Hedges G etc.) for paired samples? If pooling is valid, what is the best strategy for pooling? Is it better to pool data then calculate effect size, or calculate the effect size for each channel individually, and then average the resulting effect sizes? Or is it only valid to report the effect size for each channel separately? Thanks!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.