Cross-sectional Analysis of Longitudinal Data with Missing Values in the Dependent Variables: A Comparison of Weighted Estimating Equations with the Complete Case Analysis

Kastner, C. and Ziegler, Andreas (1997): Cross-sectional Analysis of Longitudinal Data with Missing Values in the Dependent Variables: A Comparison of Weighted Estimating Equations with the Complete Case Analysis. Collaborative Research Center 386, Discussion Paper 64

Abstract

Inference for cross-sectional models using longitudinal data can be drawn with independence estimating equations (Liang and Zeger, 1986). Many studies suffer from missing data. Robins and coworkers proposed to use weighted estimating equations (WEE) in estimating the mean structure, if missing data are present in dependent variables. In this paper the WEE are compared with complete case analyses for binary responses using simulated data. Our results are in accordance with the theoretical findings of Robins and coworkers. The WEE yield consistent estimates, even if the data are missing at random.

• Item Type: Paper (Research Paper)
• Collections: Mathematics, Computer Science and Statistics > Statistics > Collaborative Research Center 386; Special Research Fields > Special Research Field 386
• Subjects: 500 Science > 510 Mathematics
• Language: English