The empirical rule in a normal data set, where every virtual piece is divided into three deviations of the mean and the average of these 3 deviations are known as the mean. This Empirical rule was found by Sal Khan. Empirical Rule is also called as 68-95-99 Rule. Three Sigma Rule is one more name given to this rule. The deviations are divided in the following manner.

● 1st deviation
68% of the complete data in this deviation.

● 2nd deviation
95% of complete data is in 1st & 2nd deviation.

● 3rd deviation
99.7 of complete data in 1st, 2nd & 3rd deviation this means that the remaining 0.03% of data will be used for outliers that are present in every data set.

This is the reason why the empirical rule is also called as 68 – 95 – 99 rule. This also means that the center of every data set must have mean, mode and median.

Determining Standard Deviation

The Empirical Rule is specifically used to understand the data set and then you can predict the outcome. It is necessary to calculate the standard deviation first before prediction. The calculation is done in the following steps:
1. The total of the data set should be divided by the number of quantity, this is how the data set is determined.

2. The mean should be subtracted by each and every number in the set, the results should be squared.

3. The squared numbers should be used to determine the mean of each number.

4. The square roots of the means should be found.

The major data should be fall in standard deviation. Which means in between the base percentage of deviation distribution. You must not include the minor outliers.

How to use it?
Now you know that this rule is just for the prediction of the outcome of the data. After the determination of the data deviation, you can use the rule on the distribution of the data. The reason why you can predict is that you can distribute the data I set on the bases of 68 – 95 – 97 rule.

This rule should only be used where you want to predict the outcome, but remember that the empirical rule is only used where the complete data is no available. This rule is simply used to study the data and assume the same. It helps you analyse where the data will fall when the complete data is available.

Remember that if you are not able to apply the empirical rule on a specific data set that that data set is not normal and so the calculation will not be proper and the prediction outcome will be accordingly.