11:09 PM
One test for normality is used which is known as Jarque Bera Test. Under this test, we
use two dimensions.
1) Skewness
2) Kurtosis
Skewness - Symmetry of the distribution is known as Skewness. If symmetry then it will normal distribution. If Skewness is zero t indicates normal distribution. Skewness should
be zero near to zero. It means distribution is symmetrical. It may be positive and negative.
Positive Skewness = Right side Skewness
Negative Skewness = Left side Skewness
Kurtosis – Sharpness of the peak of the curve is known as Kurtosis. Statistically for
normal distribution Kurtosis should be 3.
More than 3 = flat curve
Exactly 3 = normally sharp curve
Less than 3 = very sharp curve
Jarque Bera Test depends on Skewness and Kurtosis. Jarque Bera Test denoted by “JB” and checked by chi- square test. If calculated value is less than critical value than Ho will be accepted.
No comments
Post a Comment