Never use an inappropriate analysis to make decisions. If a multilevel
model is appropriate then that is the model to use to make decisions
about the variables in the model.
Paul R. Swank, Ph.D
Professor and Director of Research
Children's Learning Institute
University of Texas Health Science Center
Houston, TX 77038
-----Original Message-----
From: SPSSX(r) Discussion [mailto:
[hidden email]] On Behalf Of
Suh-Ing Amy Hsieh
Sent: Tuesday, November 18, 2008 7:43 PM
To:
[hidden email]
Subject: Univariate analyses before multilevel models or GEE?
Hi listers:
One of aims of my dissertation is to determine the effect of hospital
hematopoietic stem cell transplant (HSCT) volume (annual number of HSCT
procedures) on health outcomes (e.g., overall survival and one-year
readmission rates) after adjusting covariates (e.g., 10 patient and 6
organizational covariates). Annual number of HSCT procedures has 5-year
data (2001 to 2005). Also, only hospitals with consecutive annual HSCT
procedures will be included in the analysis. I am planning to use
multilevel models for analyzing the main effect of hospital HSCT volume.
Before I use multilevel models or GEE to run the population-based data
with
clustered characteristics, shall I run univariate analyses (regression)
firstly for deleting those covariates without p value of < .15 to .20?
Thanks.
Amy Hsieh
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
