Écart-type

Standard deviation is a numerical value measuring how far data values are from their mean. It quantifies the spread or variation in a dataset. A small standard deviation indicates data concentrated close to the mean with slight variation, while a larger standard deviation shows data values more spread out from the mean, indicating greater variation.

Standard deviation allows comparison of consistency between datasets with identical means by measuring their spread. Two teams scoring the same average goals can be compared using standard deviation to determine which team is more consistent. The team with smaller standard deviation shows less variation in performance, making them more predictable and consistent.

Sample standard deviation, denoted by s, is used when data is drawn from a sample and uses n minus 1 in the denominator. Population standard deviation, represented by sigma, is used for entire population data and uses population size N in the denominator. This difference accounts for the additional variability inherent in sample data.

Standard deviation values are never negative; they are either positive or zero. Standard deviation equals zero only when all dataset values are identical, meaning there is no variation from the mean. In all other cases, standard deviation is positive, reflecting the spread of data around the mean.

Standard deviation shares the same units as the original dataset. If measuring goals scored, standard deviation is expressed in goals. If measuring waiting time in minutes, standard deviation is in minutes. This makes standard deviation directly interpretable in the context of the data being analyzed.

At supermarket X with two-minute standard deviation and supermarket Y with four-minute standard deviation for average five-minute wait times, supermarket Y has more variation in wait times. This means wait times at supermarket Y are more spread out from the average, while supermarket X has more consistent, predictable wait times concentrated near the mean.

Several approaches help interpret standard deviation values. The empirical method to interpret standard deviation provides practical guidelines for normal distributions. Additionally, the range rule of thumb to interpret standard deviation and Chebyshev's theorem to interpret standard deviation offer frameworks for understanding what standard deviation tells us about data distribution and concentration.