You might be ok then, depending on the quality of the
pseudorandomization. In that case, if you're worried about the possible
impact of including temporal derivatives, I'd create designs with and
without the temporal derivatives and then compare the covariance
matrices and design efficiency of the two models.
cheers,
-MH
On Mon, 2011-10-31 at 22:03 +0000, Marco Loggia wrote:
> Thank you Michael,
>
> although the events of these EVs do always occur one after another, the duration is pseudorandom...
>
> Marco
|