When analyzing data, it is common to try to give a certain order to the numbers making them visually friendlier. The most common procedure is the division by classes or categories, checking the number of individuals belonging to each class.
1. The smallest and largest value for the set is determined:
2. Define the lower bound of the first class (Li) which should be equal to or slightly lower than the lowest value of the observations.
3. Define the upper bound of the last class (Ls) that should be equal to or slightly higher than the largest value of the observations.
4. Define the number of classes (K) which will be calculated using . It must be between 5 and 20.
5. Known the number of classes defines the amplitude of each class:
6. With the knowledge of the amplitude of each class, the limits for each class (lower and upper) are defined.
Click here to see example 5
Frequency distribution is approximately symmetrical with respect to a middle class
Special case of a symmetrical distribution:
When we say that the data obeys a normal distribution, we are dealing with data that is distributed in bell form.
Frequency distribution has smaller values on one side:
Long Tails Distributions
We observed that at the extremities there is a large concentration of data compared to concentrates in the central region of the distribution.
Next: Central Trend Measurements