Date:         Sat, 8 May 1999 07:34:56 -0400
Reply-To:     "C. Robert Fenlon" <cfenlon@WPO.NCCU.EDU>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         "C. Robert Fenlon" <cfenlon@WPO.NCCU.EDU>
Subject:      determine coefficients reply
Content-Type: text/plain; charset=US-ASCII

in addition to my other comments, i meant lowering your alpha level to something lower than .05 since you have so many cases and are bound to find many significant findings. you are less likely to reject a null that is "in fact" true, hence commiting a type I error. This type of error can occur if you have a lot of cases in your data set as it seems you do. You do not want to reject a null hypothesis that could be do to large sampling error or trivial patterns.
rob fenlon