This is the program for the 2010 Joint Statistical Meetings in Vancouver, British Columbia.

Abstract Details

Activity Number: 669
Type: Contributed
Date/Time: Thursday, August 5, 2010 : 10:30 AM to 12:20 PM
Sponsor: Biopharmaceutical Section
Abstract - #306739
Title: Estimating Odds Ratio for Logistic Regression in Multicenter Trials When Data Are Sparse
Author(s): Xiaomin He*+
Companies: ICON Clinical Research
Address: 257 Baxter Drive, Phoenixville, PA, 19460,
Keywords: Odds ratio ; sparse data ; penalized maximum likelihood ; multi-center trials ; data augmentation ; logisitic regression
Abstract:

Due to the existence of complete separation or quasi-separation, the likelihood equation for a logistic regression does not always have a finite solution. Estimating odds ratio under such situation includes data augmentation methods and penalized maximum likelihood function, which provides good coverage when the true odds ratio is between 1 and 4. Simulation indicates these methods are prone to under- or overestimate if the true odds ratio goes to the extreme. In a multi-center trial where data are sparse, strata effect and possible low prevalence make them unsuitable. We present a novel method to estimate the odds ratio in logistic regression by partially augmenting data, which reduces bias significantly and has better coverage when the true odds ratio is large or tends to 0. Two studies are given to demonstrate that this new method could perform better than the existing methods.


The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.

Back to the full JSM 2010 program




2010 JSM Online Program Home

For information, contact jsm@amstat.org or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.