Normal Distribution
This iconic visual must be familiar. The name 'Bell curve' is given to it for it carries the shape of a bell. We will not be getting into the technicalities of how a Normal distribution gets built via a probability density function. If you are interested, please feel free to explore that side, but here's a simplistic explanation of the bell curve.
Say you have a data set, and you come up with a frequency distribution for the same.
If you are wondering what a frequency distribution is, please go back and visit the 'Statistical Measures' section.
Let's take a data set of the weights of the top 100 tennis players and come up with a frequency distribution. Here's how it looks.
| Weight (Kgs) | Count |
| <50 | 2 |
| 50-55 | 3 |
| 56-60 | 5 |
| 61-65 | 7 |
| 66-70 | 9 |
| 71-75 | 12 |
| 76-80 | 15 |
| 81-85 | 20 |
| 86-90 | 25 |
| 91-95 | 20 |
| 95-100 | 15 |
| 101-105 | 10 |
| 106-110 | 8 |
| 111-115 | 6 |
| 116-120 | 4 |
| 121-125 | 2 |
| >125 | 1 |
There are a few interesting aspects to this bar chart.
- The bars, though seperated by spaces, can stick to each other as this is numerical data giving the chart a bell shape. We added a red line covering the bar chart to highlight the bell shape.
- You can observe the chart is symmetrical, showing us a clean bell shape.
- In such a case as the one above, the Mean, Median and Mode are all equal, and are right at the middle of the bell curve.
Too perfect to belive, is it not? Yes, most often the data sets are not like this. Let's say 'enough of this symmetry', and look at what is more common.