# Sheppard’s Correction for Moments | ML

**Prerequisite:** Raw and Central Moments

We assume in grouped data that the frequencies are concentrated in the middle part of the class interval. This assumption does not hold true in general and **grouping error** is introduced. Such an effect can be corrected in calculating the moments by using the information on the width of the class interval.

Attention reader! Don’t stop learning now. Get hold of all the important Machine Learning Concepts with the **Machine Learning Foundation Course** at a student-friendly price and become industry ready.

**Sheppard’s Correction for grouping error** is nothing but the adjustment to calculated sample moments for the grouped data or continuous data. Prof. W.F. Sheppard proved that if the frequency distribution is continuous and the frequency tapers off to zero in both directions, the grouping error can be corrected as follows:

Let **‘c’** be the width of the class interval. Then,*Raw Moments*

*Central Moments*

**What Kind of data can be corrected?**

- This method of correction to moments is only possible for the continuous variables i.e. the continuous data.
- The width of the class interval should be equal.
- Frequencies should be symmetrical. Frequency should taper off to zero in both directions.

Consider the given distribution of marks.

Marks | Number of Students |
---|---|

0 – 10 | 1 |

10 – 20 | 6 |

20 – 30 | 11 |

30 – 40 | 17 |

40 – 50 | 21 |

50 – 60 | 16 |

60 – 70 | 13 |

70 – 80 | 7 |

80 – 90 | 5 |

90 – 100 | 2 |

For the distribution of marks above, the value for moments are given below:

**Raw Moments –**

, is the rth raw moment, where is the frequency count and is the mid value of class.

So, using the above formula for the **Raw Moment** we get following values for moments.

**Sheppard’s correction for Raw Moments –**

Similarly central moments can be corrected using **Sheppard’s Correction**.