- MNE version: 1.7.0
- operating system: macOS 12.6
I am using an openly available meg dataset. The experiment involved a 2x2 factorial design, with one factor manipulated within a block and the other across the block (condition B).
The across-block conditions always occur in the same order across all subjects.
The original study was not designed to compare data across blocks, but I intend to do so.
The data set is epoched block-wise, and a traditional baseline is applied within blocks.
As the order of the block condition is not randomized (say condition A block followed by condition B block), this results in different baselines across blocked conditions, and to adjust for these, I wanted to know whether a Regression-based baseline could be used or not as the data is already cleaned with traditional baseline correction?