Table of Contents
What Is Class Width?
Class width is the difference between the upper and lower boundaries of a class interval in a frequency distribution or histogram. Choosing the right class width affects how well the histogram reveals the shape, center, and spread of the data. Too few classes hide details; too many create noise.
The class width is calculated by dividing the range (max - min) by the desired number of classes, then typically rounding up to a convenient number.
Formulas
Guidelines for Number of Classes
| Data Points | Suggested Classes |
|---|---|
| < 25 | 5-6 |
| 25-50 | 6-8 |
| 50-100 | 7-10 |
| 100-500 | 8-15 |
| > 500 | 10-20 |
FAQ
Should I round the class width?
Yes, round up to the next convenient number. For example, a width of 8.3 should be rounded to 9 or 10, depending on the data scale. This ensures all data points are covered.
What is Sturges' rule?
Sturges' rule estimates the optimal number of classes as k = 1 + 3.322*log10(n), where n is the number of data points. It works well for roughly normal data but may underestimate classes for skewed data.