Print

Print
Hi Carmen,

Yes, the different smoothing kernels will affect the results of your analyses. See this paper: http://www.ajnr.org/content/31/4/711.full.pdf
(Henley et al 2009).
Best wishes,
Eugenia

Dr. Eugenia Radulescu, MD, PhD

Postdoctoral Research Fellow

Sackler Centre for Consciousness Science,

Brighton and Sussex Medical School, UK

On Tue, Mar 13, 2012 at 6:17 PM, Andreescu, Carmen <[log in to unmask]> wrote:
Hi All,
In a Level 2 Regression analysis I have to combine two samples – young and old adults. The young sample has been preprocessed with a smoothing kernel of 6, the old sample with a 10. If I analyze  them together in this regression model would the different smoothing kernel influence the results? Do I need to redo the preprocessing with the same smoothing kernel for both samples?
Thanks,
Carmen