A Model For Binary Response Variable With Time-Censored Observations

Abstract

The work in this thesis is concerned with the investigation and development of a new model for analyzing censored binary observations. Binary data is always expressed in the form 0 and I. In other words, the response variable can only take on the value 0 or 1, which represent the non-response or response of the variable under consideration. In the case of grouped binary data, the response is expressed in the form of proportions. In certain follow-up studies, especially during the study period, some variables may result in censored observations. This becomes a problem especially in the case of ungrouped binary observations, where 0 is assigned to the non-responding individuals, or those who have got a chance to respond but not within the study period. This gives rise to the probit-exponential model.