Modeling High-Dimensional Survey Data Using Latent Structure Analyses

• Activity Number: 664
• Type: Contributed
• Date/Time: Thursday, August 4, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Igor Akushevich*+ and Mikhail Kovtun and Anatoliy Yashin
• Companies: Duke University and Duke University and Duke University
• Keywords: Multidimensional categorical data ; demographic surveys ; latent analysis ; health state ; Grade of Membership ; Latent Class Model

Abstract

The Linear Latent Structures (LLS) analysis assumes that the mutual correlations observed in survey variables reflect a hidden property of subjects that can be described by low-dimensional random vector. The statistical properties of LLS analysis, the algorithm for parameter estimates and its implementation, simulation studies, and application of LLS model to the National Long Term Care Survey (NLTCS) data are discussed. The results of analyses are compared numerically and analytically to predictions of the Latent Class and Grade of Membership analyses. Simulation studies demonstrate high quality of reconstruction of the major model components and demonstrate its potential to analyze survey datasets with 1000 or more questions. Applying the LLS model to the 1994 and 1999 NLTCS datasets (5,000+ individuals) with responses to over 200 questions on behavior factors, functional status, and c