|
Activity Number:
|
328
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #305811 |
|
Title:
|
Missingness Screens and Regression Modeling in Clinical Aging Research
|
|
Author(s):
|
Peter H. Van Ness*+
|
|
Companies:
|
Yale University
|
|
Address:
|
1 Church Street, 7th Floor, New Haven, CT, 06510,
|
|
Keywords:
|
missing data ; regression modeling ; screening tools ; older populations
|
|
Abstract:
|
Properly handling missing data is a challenge for clinical researchers working with older populations having high levels of morbidity and mortality. Screening tools have been developed by biostatisticians that enable researchers to understand the character of missing values in a set of variables. One test assesses whether values are missing completely at random--in a way depending neither on observed nor unobserved values. Another set of indexes assesses whether the values are missing at random--in a way depending on observed values but not on unobserved values. Use of such screening tools introduces complications into variable selection. In this presentation, we describe a model fitting process that incorporates the use of missingness screens, controls for collinearity, and selects variables based on model fit. We illustrate the process in a study of ICU delirium in an older cohort.
|
- 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 2006 program |
