[R] Equation for the standard error of a predicted score for a cross-classified model

From: Jonathan Weeks <weeksjp_at_gmail.com>
Date: Thu, 13 Mar 2008 19:01:22 -0600


I have several years of longitudinal test scores for students (many who switch schools at various points in time). I am using a mixed-effects model with crossed random effects to model student trajectories. The model includes time at level 1 and students crossed with schools at level 2. When I run the model I get the posterior variances on the intercepts and slopes for students and schools, but I am trying to figure out how to combine these variance components to determine the standard error for each student's predicted score at a given point in time.

Say, for a given student

pi = posterior variance for their intercept
ps = posterior variance for their slope
si = posterior variance of the intercept for the school the student was in
at time t
ss = posterior variance of the slope for the school that the student was in at time t

This is what I'm currently thinking

SE = sqrt(pi+si+(t-x)^2(ps+ss))

where t = time and x = mean number of observations across all students.

Any help anyone can offer would be greatly appreciated.

Jonathan Weeks
Doctoral Candidate
School of Education
University of Colorado, Boulder

	[[alternative HTML version deleted]]

R-help_at_r-project.org mailing list
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Received on Fri 14 Mar 2008 - 01:04:01 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Fri 14 Mar 2008 - 01:30:21 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.

list of date sections of archive