Research articles
ScienceAsia (): 440-443 |doi:
10.2306/scienceasia1513-1874...440
An examination of the variability of agricultural production using a new approximate confidence interval for the coefficient of variation
Vorrapot Saeleea,*, Suchada Kornpetpaneea, Patcharee Wongkasemb
ABSTRACT: The objective of this study was to develop a new approximation method for determining confidence intervals for the coefficient of variation, and, using Monte Carlo simulation, to compare the relative efficiency of the new method with Miller's, Vangel's, and two types of Mahmoudvand and Hassani's methods. It was found that the efficiency of the new approximation method was superior when n≥30 and the population coefficient was greater than 0.5. With n≥30 and the population coefficient less than 0.5, the new method was no less efficient than the other methods, and sometimes better. The new method was applied to examine the variability of a variety of agricultural products using data from 2006–2010. It was found that the variation in cassava yields was less than that for maize, rubber, and main rice. Cassava production was more stable across all environments.
Download PDF
7 Downloads 1361 Views
| a |
Research and Statistics in Cognitive Science, Burapha University, 169 Longhad Road, Bangsaen, Chonburi 20131 Thailand |
| b |
Faculty of Science, Burapha University, Thailand |
* Corresponding author, E-mail: vorrapot@gmail.com
Received 14 Mar 2012, Accepted 1 Feb 2013
|
