A Bayesian Approach for Estimation of Coefficients of Variation of Normal Distributions

The coefficient of variation is widely used as a measure data precision. Confidence intervals for single when the follow normal distribution that symmetrical and difference between coefficients two populations are considered in this paper. First, confidence obtained with adjusted generalized interval (adjusted GCI), computational, Bayesian, Bayesian approaches. These approaches compared existing ones comprising approximately unbiased estimators, method variance estimates recovery (MOVER) (GCI). Second, distributions proposed using same approaches, performances which then highest posterior density was to estimate interval. Monte Carlo simulation assess performance intervals. results studies demonstrate were more accurate better than others terms coverage probabilities average lengths both scenarios. Finally, all scenarios illustrated via an empirical study real-data examples.