Comparing Distributions with Different Standard Deviations

Two data sets can share the same mean and still look nothing alike, because the mean only tells you where the center is, not how the values sit around it. Standard deviation fills that gap: it measures the typical distance between each value and the mean. A set where every point sits right on the mean has a standard deviation of zero, no matter what that mean is. A set where points are scattered far from the mean has a large standard deviation, even if its mean matches the first set exactly.

You can often compare two dot plots by eye. Look at each set's mean, then ask which set has points sitting farther away from that mean on average. A tight cluster near the center means low spread. Points scattered across a wide interval mean high spread. Take 10, 10, 10, 10 against 4, 8, 12, 16: both have mean 10, but the first set never leaves 10 while the second set spreads out to 4 and 16. No arithmetic needed to see the second has the larger standard deviation.

This comparison skill connects to how the test treats spread alongside other distribution ideas, like shape and outliers. One scaling rule is worth knowing cold: multiply every value in a set by some factor k, and the standard deviation also multiplies by k, since every distance from the mean stretches by that same factor. Adding the same number to every value, by contrast, shifts the mean but leaves the standard deviation unchanged, since the distances between points don't change.