A Comparison of Knot Selection Algorithms for B-Spline Varying--Coefficient Methods When Controlling for Nonignorable Dropout

• Activity Number: 145
• Type: Contributed
• Date/Time: Monday, August 5, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Camille Moore*+ and Samantha MaWhinney and Jeri Forster and Nichole Carlson
• Companies: Colorado School of Public Health, University of Colorado Denver and Colorado School of Public Health and Colorado School of Public Health and Colorado School of Public Health, University of Colorado Denver
• Keywords: Dropout ; Missing data ; Varying Coefficient Models

Abstract

Dropout is a common source of missing data in longitudinal studies, and often occurs for reasons that may be related to an outcome of interest. When dropout depends on unobserved outcomes, even after conditioning on observable data, data are missing not at random and are therefore not ignorable. When the dropout mechanism is unspecified, semiparametric varying coefficient models can be used to account for nonignorable dropout. This method is more robust than the parametric conditional linear approach. However, fitting b-spline varying coefficient models requires the specification of the number and location of spline knots, which may influence model fit. We present simulation results comparing the natural cubic B spline varying coefficient method with a knot location selection algorithm to the same method with knots evenly placed at the quantiles of the dropout distribution, as well as to the classic conditional linear model.