WEIGHTED LEAST-SQUARES ANALYSIS OF REPEATED CATEGORICAL MEASUREMENTS WITH OUTCOMES SUBJECT TO NONRESPONSE

In this paper, we describe a two-step weighted least squares method fo r analyzing repeated categorical outcomes when some individuals are no t observed at all times of follow-up. Other weighted least squares met hods for analyzing repeated measures data with missing responses have previously been proposed by Koch, Imrey, and Reinfurt (1972, Biometric s 28, 663-692) and Woolson and Clarke (1984, Journal of the Royal Stat istical Society, Series A 147, 87-99). These methods give consistent e stimators if the responses are missing completely at random, as discus sed in Rubin (1976, Biometrika 63, 581-592). We propose a two-step met hod that will give consistent results under the weaker condition of mi ssing at random, and compare it with the other two methods.