ASYMMETRY OR POINTINGMEASUREMENTS

ASYMMETRY OR POINTINGMEASUREMENTS

1. Its concept lies in the fact that when the measures do not coincide they are used to to check whether a statistical distribution can be accepted as following the normal distribution.

2. Measures (kurtosis)

   • If g * 2(X) > 0 , the distribution is leptokurtic, more pointed than the Normal distribution.
   • g * 2(X) = 0 , the distribution is mesocurtic, equally sharper than Normal.
   • g * 2(X) < 0 , the distribution is platykurtic, more flattened than Normal.

3. Measures of skewness (kurtosis)

   • They analyze whether a frequency distribution is more or less pointed by comparing it with a standard distribution, the Normal distribution.

4. Skewness measures

   • To measure asymmetry it is common to obtain the so-called asymmetry coefficient (Fisher's), which is denoted by g1 and is defined as the quotient between the central moment of order 3 and the standard deviation raised to the cube.

5. Classification of asymmetry

   • Negative asymmetric
     • when the most frequent values of the variable are the largest and the distribution has a left tail.
   • Positive asymmetric.
     • when the most common values of the distribution are the smallest, tail to the right.

Bilder und Dateien der Mindmap