Sensitivity Analysis of Longitudinal Data with Intermittent Missing Values: Application in a Clinical Trial | ||||
| The Egyptian Statistical Journal | ||||
| Article 1, Volume 61, Issue 1, June 2017, Page 1-14 PDF (4.44 MB) | ||||
| Document Type: Original Article | ||||
| DOI: 10.21608/esju.2017.270055 | ||||
 View on SCiNiTO | ||||
| Author | ||||
| Abeer S. Ahmed* | ||||
| The National Centre for Social and Criminological Research, Cairo, Egypt | ||||
| Abstract | ||||
| We conduct two types of sensitivity analyses of study conclusion , in intermittent setting : the first is fitting different to the response variable. The marginal distribution of the response is assumed to be an extreme valuea logistic distribution, the lognormal distribution in particular. The second is fitting several models for the missing data mechanism, for example, modeling the missingness based on a generalized linear model, with logic and probit link function. The model can be extended to permit possible relationships between the missing data process and covariates, for example time. The selection model for incomplete longitudinal data is presented. InternatioThe stochastic EM algorithm is proposed and developed for skewed distribution model, the lognormal distribution in particular. Models for the missing data mechanism are presented. The proposed methods are applied to a data set from the nal Breast Cancer Study Group. | ||||
| Keywords | ||||
| Longitudinal Data; Missing values; Clinical Trial | ||||
|
Statistics Article View: 93 PDF Download: 258 |
||||
