Computational Statistics & Data Analysis
The nature of sensitivity in monotone missing not at random models
Computational Statistics & Data Analysis
Computational Statistics & Data Analysis
| Hi-index | 0.03 |
Diggle and Kenward (Appl. Statist. 43 (1994) 49) proposed a selection model for continuous longitudinal data subject to possible non-random dropout. It has provoked a large debate about the role for such models. The original enthusiasm was followed by skepticism about the strong but untestable assumption upon which this type of models invariably rests. Since then, the view has emerged that these models should ideally be made part of a sensitivity analysis. One of their examples is a set of data on mastitis in dairy cattle, about which they concluded that the dropout process was non-random. The same data were used in Kenward (Statist. Med. 17 (1998) 2723), who performed an informal sensitivity analysis. It thus presents an interesting opportunity for a formal sensitivity assessment, as proposed by Verbeke et al. (sensitivity analysis for non-random dropout: a local influence approach, 2000; submitted), based on local influence (Cook, J. Roy. Statist. Soc. Ser. B 48 (1986) 133).