Hi Andre,
you are perfectly right, see this post from Jesper Andersson:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind06&L=SPM&P=R733457&1=SPM&9=A&I=-3&J=on&d=No+Match%3BMatch%3BMatches&z=4
Best,
Guillaume
Le 28/07/2010 10:56, Andre Szameitat a écrit :
> Dear SPMers,
> I was just wondering whether a parametric manipulation with 3 levels of
> the manipulated variable (e.g. low, medium, high) makes any sense.
> AFAIK, SPM modifies each regressor so that its mean is zero. Thus, if I
> code a 3 level parametric manipulation as 1, 2, 3, SPM changes this to
> -1, 0, 1 - this actually means that the medium level is ignored. Is this
> correct, or am I missing something?
>
> Many thanks&
> Best Regards,
> Andre
>
|