EMPIRICAL COMPARISON OF RELATIVE PRECISION OF GEOMETRIC MEASURE OF VARIATION ABOUT THE MEAN AND STANDARD DEVIATION

Benedict Troon

No tzkw9, AfricArxiv from Center for Open Science

Abstract: Measure of dispersion is an important statistical tool used to illustrate the distribution of datasets. The use of this measure has allowed researchers to define the distribution of various datasets especially the measures of dispersion from the mean. Researchers have been able to develop measures of dispersion from the mean such as mean deviation, mean absolute deviation, variance and standard deviation. Studies have shown that standard deviation is currently the most efficient measure of variation about the mean and the most popularly used measure of variation about the mean around the world because of its fewer shortcomings. However, studies have also established that standard deviation is not 100% efficient because the measure is affected by outlier in the datasets and it also assumes symmetry of datasets when estimating the average deviation about the mean a factor that makes it to be responsive to skewed datasets hence giving results which are biased for such datasets. The aim of this study is to make a comparative analysis of the precision of the geometric measure of variation and standard deviation in estimating the average variation about the mean for various datasets. The study used paired t-test to test the difference in estimates given by the two measures and four measures of efficiency (coefficient of variation, relative efficiency, mean squared error and bias) to assess the efficiency of the measure. The results determined that the estimates of geometric measure were significantly smaller than those of standard deviation and that the geometric measure was more efficient in estimating the average deviation for geometric, skewed and peaked datasets. In conclusion, the geometric measure was not affected by outliers and skewed datasets, hence it was more precise than standard deviation.

Date: 2021-03-25
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://osf.io/download/605b67347d552b00ee1b92e1/

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:osf:africa:tzkw9

DOI: 10.31219/osf.io/tzkw9

Access Statistics for this paper

More papers in AfricArxiv from Center for Open Science
Bibliographic data for series maintained by OSF ().