[cc'ing back to r-sig-mixed-models]
On 04/14/2011 11:07 PM, Sujal Phadke wrote:
Thank you, now I understood everything except how the number of levels affects our decision of randomness of an effect? As it is probably clear by now, I have hardly dealt with advanced stats or mixed models so please bear with me for such questions. I would appreciate your help. Thanks Sujal
See <http://glmm.wikidot.com/faq#fixed_vs_random> for a discussion of this point, especially the latter half of the section and the long mailing-list thread that is pointed to there ... good luck Ben Bolker