variance

Variance

The average squared deviation from the mean — the formal measure of how spread out data values are around their center.

When you'd see it: In ANOVA (analysis of variance), in regression model diagnostics, and in ML model discussions (the bias-variance tradeoff). Standard deviation is more commonly reported in practice because it's in the same units as the data, but variance is what the math operates on.

Why it matters: Variance is the formal building block for most inferential statistics. In ML, high variance means a model that overfit the training data — it performs well in-sample but poorly on new data. Understanding variance is the prerequisite for understanding the bias-variance tradeoff.

Common mistakes: Confusing variance with standard deviation in reported numbers. Variance is in squared units — a variance of 400 in a salary dataset means a standard deviation of $20, not $400.

Related standard deviation, mean, ANOVA, overfitting, descriptive statistics