Hi Philipp, Susie,
> Approach 1 indeed disregards the higher power of the parametric statistics,
> and correlation within the patient group between SPMs and performance
> variables seems more advantageous.
It indeed does seem so. However, depending on how broad a measure of
cognitive ability you assess, be mindful of the correlation with global
tissue volume. When we looked at IQ a while ago, we needed to divide
subjects into broader groups as otherwise the covariate of interest
correlated with the globals, for which we wanted to correct. This helped
us to find the "tip of the iceberg", at least.
> In any case, it is advisable to check correlation plots, as
> also in SPM outliers can easily produce spurious effects.
That is most certainly a good advice :)
Marko
--
=====================================================================
Marko Wilke (Dr.med./M.D.)
[log in to unmask]
Universitäts-Kinderklinik University Children's Hospital
Abt. III (Neuropädiatrie) Dept. III (Pediatric neurology)
Hoppe-Seyler-Str. 1, D - 72076 Tübingen
Tel.: (+49) 07071 29-83416 Fax: (+49) 07071 29-5473
=====================================================================
|