[Eeglablist] Doubts

[Filipa Câmara]

Hi everybody,
I am currently doing my master thesis and investigating perception and execution processes, with a particular focus on comparing autistic and neurotypical individuals. This research emphasizes the role of mirror neurons and their relationship with the mu rhythm. The study includes two paradigms—dance and walking—with three tasks administered in each paradigm (as it can be seen on the images).
I created a Study in EEGLAB by integrating all the datasets. During the pre-computation of measures, I used the default parameters for ERSPs (this is the main measure that I will use in my study)—wavelet cycles set to [3 0.8], nfreqs = 100, and ntimesout = 200—without explicitly specifying the baseline, because it was not there as an "input" parameter. According to the documentation, I understood that the default is baseline = 0, meaning that the pre-stimulus period (from -1000 to 0 ms) is automatically assumed as the baseline.
The ERSP plot for the instruction phase (0–3000 ms) appears as expected, while the ERSP for the execution phase (3500–7500 ms) exhibits a pattern that does not align with what is observed in the same time interval in the full-epoch plot (first image). I've used a frequency range of 4-40Hz for all three images and I was setting diferent scales (power limits) for better visualization on "Entire time ERSP plot" or "Instruction/Execution time" Ersp plot.
I would like to clarify the following questions:

  1.
Considering that I did not explicitly specify the baseline during pre-computation—relying on the default baseline = 0 (i.e., using the -1000 to 0 ms interval as the baseline)—why does the ERSP generated for the execution sub-interval (3500–7500 ms) show a pattern that differs from that observed in the corresponding interval of the full-epoch plot? Might this be due to a recalculation of the baseline or edge effects related to the winsize?
  2.
What can be the problem of the "execution epoch" that seems that it does not fit on the pattern of the full epoch?

Thank you very much for your time and for any guidance you can provide, as these questions are crucial for correctly interpreting my data.
Sincerely,