Hi Federico,
> How can I fix this trade-off problem? How do you suggest me to compute
> the high pass cut-off? Is it possible to get the same signal losses
> for all experimental conditions when the order is randomized?
You have 6 conditions, but these might (hopefully) be alternatively
expressed as a 2x3 factorial design. The contrasts between levels of
either factor, or the interaction, would have much shorter periodicity
than the contrasts expressing individual conditions. By reformulating
your design in terms of contrasts between levels of each factor, =/- the
interaction, you will require a much shorter HPF, therefore less noise,
without loss of signal for the effects you are interested in.
James
|