###### Asked in Home & GardenMathematical FinanceStatisticsAlgebra

# What is the use of coefficient of variability in statistical analysis in explaining analytical result?

## Answer

###### Wiki User

###### November 01, 2007 6:13PM

The coefficient of variablility is usually referred to as
R-squared. It is the percentage (written in decimal form - like .80
means 80%) of the variance in the data that is explained. You want
that number to get as close to 1.00 (which means 100%) as possible.
If your R-squared is .65, that means that you have explained 65% of
the variance, or fluctuation, in the data. To get the percentage
higher, you can add more variables to your model, or attempt
transformations of the current variables in your model. There is no
set value that the R-squared needs to be - it is dependent on what
type of analysis you are doing and what you are trying to explain.
Be cautious in adding additional variables to a model just to make
only a small gain in your R-squared (like 2% or less), as more
variables means more potential for multicollinearity in your model.
The Coefficient of Variability (CV) allows comparison of the
standard deviations of different variables that are in different
units of measure. For example, if you wanted to compare the length
of a course of recovery from a specific infectious illness with the
number of times that the patient had had that illness, you could
approach the study with the coefficient of variability. In that
instance, a CV might tell you -- if the numbers happened to work
out this way -- that in your sampled population, relative to their
means, the variability in length of illness was greater than the
variability in number of times the patients had had the illness.
Technically, CV is used with *ratio scale variables* where
zero is an "absolute" zero point; i.e. a score of 0 = nothing.

CV = [(100) (s) / X ]

This statistic measures the ratio of the standard deviation of a variable relative to its mean