The new estimation of coeffi cient of variation of ratio correlated data

Abstract

The objectives of this study are to propose the new estimation coeffi cient of variation method of ratio
correlated data and to compare with original coeffi cient of variation method. The research was considered by
Root Mean Square Error (RMSE). The comparisons were done by using sample sizes (n) equal to 5, 10, 15, 30, 50,
and 100 whereas coeffi cient correlation values (ρ) are ±0.3, ±0.5, ±0.7 and ±0.9. This study used the Monte Carlo
Simulation method and the experiment was repeated 1000 times for each condition. The results showed that the
new methods to estimate the coeffi cient variation generates the RMSE value lower than the original methods in
every condition.