|
Activity Number:
|
324
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
ENAR
|
| Abstract - #309495 |
|
Title:
|
A Comparison of Various Analytical Methods for U-Shape Distributed Random Variables: A Real-World Application from a Behavioral Intervention Trial
|
|
Author(s):
|
Terry Wang*+ and Torrance T. Stephens and Lillian S. Lin
|
|
Companies:
|
Centers for Disease Control and Prevention and Morehouse School of Medicine and Centers for Disease Control and Prevention
|
|
Address:
|
1600 Clifton Rd NE MS E48, Atlanta, GA, 30333,
|
|
Keywords:
|
cross section ; proportional odds ; survey ; multivariate
|
|
Abstract:
|
HIV behavioral intervention trials target risky sexual behaviors for different at-risk populations. Various measures are used to assess changes in risky sexual behaviors, particularly condom use during sex, among the target populations. One construct typically used for analyses is the proportion of times a subject reports using a condom during a specific period. This random variable often follows a U-shaped distribution; standard linear regression procedures may not be valid. Other methods for operationalizing the outcome are also considered. We will present our findings from research studies based on the various operationalizations of condom use and will compare and contrast the various analytical approaches and their interpretations.
|
