|
Activity Number:
|
209
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Social Statistics Section
|
| Abstract - #309998 |
|
Title:
|
Nonlinear Random Coefficients Regression for Estimating an Interaction Term
|
|
Author(s):
|
Reid Landes and Jeffrey A. Pitcock*+
|
|
Companies:
|
University of Arkansas for Medical Sciences and University of Arkansas for Medical Sciences
|
|
Address:
|
4301 W Markham St, Little Rock, AR, 72205,
|
|
Keywords:
|
delay discounting ; impulsivity ; clinical trial ; heterogeneity of variance
|
|
Abstract:
|
Many clinical trials are designed to compare how much a standard and experimental population change between pre- and post-treatment. In other words, the interaction of treatment and time---a difference of differences---is of interest. We consider how to make this comparison when an individual's outcome of interest is estimated with nonlinear regression. In particular, we describe how to use all of the raw data in a nonlinear random coefficients regression that allows direct estimation of the difference of differences. Further, we also illustrate how to account for heterogeneity of variance among the individuals. We apply this method to impulsivity data collected from treatment-seeking opiate addicts in a two-treatment clinical trial. The results from the full nonlinear regression are compared with ad hoc methods which are typically used for these sorts of problems.
|
