|
Activity Number:
|
197
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 5, 2003 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #300674 |
|
Title:
|
Indications and Limitations of Structural Equation Modeling in Complex Surveys: Implications for an Application in the Medical Expenditure Panel Survey
|
|
Author(s):
|
Janet L. Valluzzi*+ and G. Edward Miller and Sharon L. Larson
|
|
Companies:
|
Agency for Healthcare Research & Quality and Agency for Healthcare Research & Quality and Agency for Healthcare Research & Quality
|
|
Address:
|
2101 East Jefferson St., Rockville, MD, 20852,
|
|
Keywords:
|
structural equation modeling ; complex survey design ; analytical techniques
|
|
Abstract:
|
Structural equation modeling (SEM) is increasingly used as an analytical option that focuses on latent or unobserved variables in model confirmation. This paper examines types of research questions that can be addressed with SEM and the underlying assumptions and limitations of SEM techniques. Dataset requirements, particularly those related to complex multistage probability sample designs, are identified; distinguishing between issues related to stratification and unequal inclusion probabilities and issues related to clustering. While there are a number of suggestions in the literature that discuss options for use of SEM; there is little demonstration in the literature of its application in complex survey data analysis. The Medical Expenditure Panel Survey (MEPS), a complex multistage probability survey conducted annually since 1994, estimates health care utilization and expenditures for the U.S. civilian noninstitutionalized population. Using MEPS data, this paper discusses an application of SEM to examine national estimates of prescription medication utilization and expenditures among children and youth who use psychotropic medications.
|
