Date:         Fri, 27 Jan 2012 11:22:30 -0500
Reply-To:     William Shakespeare <shakespeare_1040@HOTMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         William Shakespeare <shakespeare_1040@HOTMAIL.COM>
Subject:      Mixed model

Say I have students nested in classrooms and I've measured a set of
predictors and an outcome.  Furthermore, say I incorrectly ignore the
clustering and run a linear model rather than a mixed model.  Are the
standard error on the predictors coefficients underestimated or
overestimated?