If the frequency of observations has equal distance from an average are the same, then we say that distribution is summetric, that is mean=median=mode.
If the frequency distribution is not symmetric then we say that distribution is asymmetric or skewed.
SKEWNESS
The skewness is the unsymmetric of a frequency distribution. There are two possibilities for a distribution is skewed that is ‘positively skewed ‘ and ‘negatively skewed ‘.
1) Postively Skewed :
The frequency curve has a longer tail on the right of the mode, it is prove that is a positively skewed distribution. Mean greater than median.
Mean > Median > Mode
2) Negatively Skewed :
The frequency curve has the longer tail on the left of the mode.It is prove that it is in a negatively skewed distribution. That is
Mean < Median < Mode.
Different of co-efficient of skewness :
1) Karl peason’s co-efficient
of skewness
a) Mean – Mode
—————————–
SD
b) 3(Mean – Median )
——————————
SD
[SD= Standard Deviation ]
2) Bowley’s co-efficient of
skewness.
(Q3 + Q1) – (2Median)
Q3 - Q1
KURTOSIS
It refers to the peakness of a frequency distribution. Using the value of beta two the frequency distribution can be divided into 3.They are,
1) Lepto Kurtic distribution
It is a frequency distribution for which beta two grater than 3.
2) Meso Kurtic distribution /
Normal distribution.
It is a frequency distribution for which beta two equal to 3.
3) Platy Kurtic distribution
It is the frequency distribution for which beta two less than 3.