JSM 2013 Home
Online Program Home
My Program

Abstract Details

Activity Number: 421
Type: Topic Contributed
Date/Time: Tuesday, August 6, 2013 : 2:00 PM to 3:50 PM
Sponsor: Biometrics Section
Abstract - #308712
Title: Evaluation of Approaches to Analyzing Clustered Data When the Number of Clusters and Cluster Size Are Small: A Simulation Study
Author(s): Jiayan Huang*+ and Gui-shuang Ying
Companies: University of Pennsylvania and University of Pennsylvania
Keywords: clustered data ; generalized estimating equations (GEE) ; Wald test ; Score test ; paired t-test ; linear mixed model (LMM)
Abstract:

Clustered data arise from measurements in a small number of subjects over time or at multiple sites. Little is known about the relative performance of available statistical methods when both number of clusters (N) and units per cluster are small. Motivated by vision research, we simulated data under 2 designs. In design 1, two eyes of a subject are in 2 different groups; in design 2, eyes are in the same group. Under different inter-eye correlation (r), N (5 to 50), and effect size, we evaluated Wald and score tests from generalized estimating equations (GEE), F-test from linear mixed model (LMM), paired t-test (design 1), and the 2 sample t-test using the mean of 2 eyes or a randomly chosen eye (design 2). Type I error rate (alpha) and statistical power were the performance criteria. In design 1, the paired t-test and LMM perform best with nominal alpha, and higher power. In design 2, no test performed uniformly well. The t-test has nominal alpha but lower power. LMM has inflated alpha in design 2. In both designs, the GEE Wald test inflates alpha and the score test has lower alpha and power, consistent with other studies. The performance of each test is dependent on the r and N.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2013 program




2013 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.

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

ASA Meetings Department  •  732 North Washington Street, Alexandria, VA 22314  •  (703) 684-1221  •  meetings@amstat.org
Copyright © American Statistical Association.