範囲

The range is a measure of variation defined as the difference between the maximum and minimum values in a dataset. To calculate it, subtract the smallest value from the largest value. For example, in a dataset of soda can volumes, if the highest measurement is 16.1 ounces and the lowest is 14.8 ounces, the range is 16.1 minus 14.8, which equals 1.3 ounces. This simple calculation makes the range easy to compute quickly.

The range relies only on the maximum and minimum values, ignoring all other data points. This means a single extreme outlier can dramatically change the range value, even if most data points cluster closely together. Because the range depends entirely on these two extreme values, it lacks robustness and may not accurately represent the typical variation within a dataset.

The range is widely used in statistical process control and manufacturing quality assurance because it is easier to compute than other measures of variation. Manufacturers regularly use the range to monitor whether product measurements, such as soda can volumes, fall within desired specifications. Weather forecasting also employs the range to track temperature and precipitation variations, making it valuable for real-time monitoring applications.

Unlike other measures of variation such as standard deviation or mean absolute deviation, the range uses only extreme values rather than considering the entire dataset. While this makes the range quick to calculate, it provides less information about how data is distributed. The range rule of thumb to interpret standard deviation offers a more comprehensive approach to understanding data spread and variability.

In a dataset comparing dog breed weights, the range represents the weight difference between the heaviest breed, the English Mastiff, and the lightest breed, the Chihuahua. A range of 101 kilograms indicates the total weight variation among different dog breeds. This tells us that any individual breed's average weight will fall somewhere within this 101-kilogram spread or dispersion.

Measurements of soda can volume vary because different people record measurements and the exact amount of liquid poured into containers may not be precisely 16 ounces. These variations create a range of values around the target volume. Manufacturers test regularly to ensure the actual amount of soda falls within an acceptable range, accounting for these natural measurement and filling inconsistencies.

The range has limited usefulness because it is highly susceptible to outliers and lacks robustness in measurement. A single unusually high or low value can inflate or deflate the range, misrepresenting the typical data spread. Despite these limitations, the range remains popular in manufacturing and weather forecasting because it is simple to compute and provides quick estimates of data dispersion.