Estimation for Censored Exponential Data When the Censoring Times are Subject to Error
M. J. Phillips, T. J. Sweeting
Journal of the Royal Statistical Society: Series B (Statistical Methodology) · 1996 · Volume 58 · Issue 4
DOI: 10.1111/j.2517-6161.1996.tb02115.x
Abstract
The problem of estimating the parameter of the negative exponential distribution with censored data may be complicated by a lack of complete knowledge of the censoring times. The effect of ignoring these errors in the censoring times rather than using a complete maximum likelihood estimation approach is investigated and it is shown that the loss in efficiency is generally small for the inverted gamma model considered in this paper. The theory is supplemented by some numerical results based on an example of failure data.
Keywords
censored data, errors in observation, inverted gamma distribution, maximum likelihood estimation, negative exponential distribution
Enable JavaScript for citation tools and additional paper interactions.