Hi,
There's no hard rule on this - even your "10" isn't necessarily a hard
limit. As you include more and more confound regressors you lose
degrees of freedom and start to get correlations maybe with the model,
both of which can reduce estimation efficiency, but there's no fixed
limit.
The statistical modelling will always adjust so that you still get
valid stats even in the cases where you're losing efficiency.
Cheers.
On 10 Apr 2009, at 13:41, Mahinda Yogarajah wrote:
> Hi Steve,
>
> Thank you for such a prompt reply.
>
> Another quick question to the list - is there a limit to the number of
> confounding factors one can include in a correlational TBSS
> analysis, as
> there would be if one was simply carrying out an ordinary statistical
> partial correlation analsyis or a linear regression (where one might
> assume
> that one would need at least 10 subjects for every predictor in the
> model to
> prevent danger of "over modelling").
>
> Thanks.
>
> Mahinda
>
>
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|