Abstract:
|
We propose a three-parameter distribution called the reflected shifted truncated gamma (RSTG) distribution that can be used to model left skewed data. We provide explicit expressions for the mean, variance, hazard function, survival function and quantile function. We estimate the model parameters by maximum likelihood methods based on complete, right censored and interval censored survival data. We assess the performance and verify the consistency of the maximum likelihood estimators of the RSTG by conducting a simulation study with varying sample sizes. Finally, we compare the RSTG distribution to the exponential, Generalized F, Generalized gamma, Gompertz, log-logistic, lognormal, Rayleigh, and Weibull distributions for modeling left skewed data in three data sets. We compare them using AIC, AICC, HQIC, and CAIC criteria and conclude that the RSTG distribution is superior to other distributions in modeling left skewed complete and right censored survival data.
|