Analysis of Familial Aggregation with Complex Survey Data

• Activity Number: 462 - SPEED: Survey Research Methods
• Type: Contributed
• Date/Time: Wednesday, August 2, 2017 : 8:30 AM to 10:20 AM
• Sponsor: Survey Research Methods Section
• Author(s): Cong Wang* and Zhaohai Li and Barry I. Graubard
• Companies: George Washington University and George Washington University and National Cancer Institute
• Keywords: familial aggregation ; network sampling ; recurrence risk ; propensity score weighting ; varying family sizes ; different family relationships

Abstract

Familial aggregation is considered as an important aetiology of disease so that many studies focus on the analysis of familial aggregation. In recent studies, marginalization approach is considered as a better method to analyze the familiar aggregation of varying family sizes. The purpose of our research is to combine the marginalization approach with complex survey design, analyzing the familial aggregation for the families with different familial relationships and varying family sizes. Weights are added to recent model to obtain the parameter estimators and the robust variance estimators. The recurrence risk is what we use to represent the familial aggregation. We apply our model to diabetes disease data collected by the National Center for Health Statistics (NCHS), Centers for Disease Control and Prevention (CDC) in 1976. Network sampling method is used and propensity score weighting is also applied to adjust for the compounding variables. Also, simulation studies are conducted to examine the parameter estimates and variance estimators in the weighted model.