For a distribution, the percentile partition values are P 10 = 58.983, P 50 = 61.345 and P 90 = 63.831. Kelly's coefficient of skewness is:
0.026
Kelly's coefficient of skewness is a measure used in statistics to describe the asymmetry of a probability distribution. Unlike Pearson's coefficient, which uses mean and standard deviation, Kelly's coefficient uses percentiles (or deciles). It is particularly useful when the data is not symmetrical or contains outliers, as percentiles are less sensitive to extreme values than the mean.
The coefficient helps us understand if the data is skewed to the left (negatively skewed), skewed to the right (positively skewed), or symmetric. A positive coefficient indicates positive skewness (tail on the right), a negative coefficient indicates negative skewness (tail on the left), and a coefficient near zero suggests symmetry.
Kelly's coefficient of skewness (Sk) is calculated using the 10th, 50th, and 90th percentiles. The formula is given by:
$$ \text{Sk} = \frac{P_{90} + P_{10} - 2P_{50}}{P_{90} - P_{10}} $$
Where:
We are provided with the following percentile partition values for the distribution:
Now, let's substitute the given values into Kelly's coefficient formula and calculate the skewness.
Step 1: Calculate the numerator
The numerator is $P_{90} + P_{10} - 2P_{50}$.
$$ \text{Numerator} = 63.831 + 58.983 - 2 \times 61.345 $$ $$ \text{Numerator} = 122.814 - 122.690 $$ $$ \text{Numerator} = 0.124 $$
Step 2: Calculate the denominator
The denominator is $P_{90} - P_{10}$.
$$ \text{Denominator} = 63.831 - 58.983 $$ $$ \text{Denominator} = 4.848 $$
Step 3: Calculate Kelly's Coefficient
Now, divide the numerator by the denominator.
$$ \text{Sk} = \frac{\text{Numerator}}{\text{Denominator}} $$ $$ \text{Sk} = \frac{0.124}{4.848} $$ $$ \text{Sk} \approx 0.0255735561 $$
Rounding the result to three decimal places, we get:
$$ \text{Sk} \approx 0.026 $$
The calculated value for Kelly's coefficient of skewness is approximately 0.026. This value is positive and close to zero, suggesting that the distribution is slightly positively skewed. The right tail of the distribution is slightly longer or fatter than the left tail.
| Measure | Formula (common) | Based On | Sensitivity to Outliers |
|---|---|---|---|
| Kelly's Coefficient of Skewness | $\frac{P_{90} + P_{10} - 2P_{50}}{P_{90} - P_{10}}$ | Percentiles (Deciles) | Less Sensitive |
| Pearson's First Coefficient | $\frac{\text{Mean} - \text{Mode}}{\text{Standard Deviation}}$ | Mean, Mode, Standard Deviation | More Sensitive |
| Pearson's Second Coefficient | $\frac{3(\text{Mean} - \text{Median})}{\text{Standard Deviation}}$ | Mean, Median, Standard Deviation | More Sensitive |
Percentiles are values that divide a dataset into 100 equal parts. The nth percentile is the value below which n percent of the observations fall. Key percentiles include:
The range between $P_{10}$ and $P_{90}$ covers the central 80% of the data. Kelly's coefficient uses these specific percentiles to avoid being heavily influenced by the extreme ends of the distribution, which are more susceptible to outliers. This makes it a robust measure of skewness.
If the data are skewed, which option of central tendency measure is the most unreliable indicator?
In a negatively skewed distribution
If the distribution is negatively skewed, then the:
The first four moments about the mean of distribution are 0, μ 2, 0.7 and 18.75. If the distribution is mesokurtic, the value of μ 2, is
If Mean > Median > Mode, the distribution is: