Calculate range, variance, standard deviation, IQR, MAD, CV, skewness, and kurtosis from any dataset with deviation tables, visual spread plot, and outlier detection.
Measures of dispersion describe how spread out data values are around the center. While the mean tells you the typical value, dispersion reveals whether values cluster tightly or scatter widely. Two datasets with the same mean can have wildly different spreads — and understanding this spread is essential for statistical inference, quality control, and decision-making.
This calculator computes every major dispersion measure from your data: range, sample and population variance, sample and population standard deviation, interquartile range (IQR), mean absolute deviation (MAD), median absolute deviation, coefficient of variation (CV), standard error, skewness, excess kurtosis, and the dispersion index (VMR). It also detects outliers using the 1.5×IQR rule.
The deviation table breaks down every data point's contribution to the variance calculation, showing deviation from the mean, absolute deviation, and squared deviation. The visual spread plot displays data points on a number line with mean, IQR box, and outlier highlighting. A comparison table explains when to use each measure, helping you choose the right statistic for your context.
Understanding data spread is just as important as understanding the center. This calculator provides a comprehensive dispersion analysis in one tool — no need to calculate range, variance, IQR, and outliers separately. The deviation table shows exactly how variance is computed step by step, making it an ideal learning tool.
The visual spread plot and measure comparison table provide insights that raw numbers alone cannot. Seeing data points with outliers highlighted and the IQR box overlay gives immediate intuition about distribution shape. The comparison table helps you choose the right measure for your specific context — a question students and professionals alike frequently face.
Range = max − min Sample Variance: s² = Σ(xᵢ − x̄)² / (n − 1) Sample SD: s = √s² IQR = Q3 − Q1 MAD = Σ|xᵢ − x̄| / n CV = (s / |x̄|) × 100% SE = s / √n Skewness = [Σ(xᵢ − x̄)³/n] / σ³ Kurtosis = [Σ(xᵢ − x̄)⁴/n] / σ⁴
Result: Range: 30, s² = 93.07, s = 9.65, IQR = 16.5, CV = 12.25%
The test scores span from 65 to 95 (range 30). The standard deviation of 9.65 means most scores fall within about 10 points of the mean (80.7). The CV of 12.25% indicates moderate relative variability. IQR of 16.5 captures the middle 50% of scores.
**Range** is the simplest but least informative — it uses only two values and is extremely sensitive to outliers. Use it for a quick first look. **Variance and standard deviation** are the workhorses of statistics, required for confidence intervals, hypothesis tests, and regression. However, they give outsized weight to extreme values (because deviations are squared).
**IQR** is robust: it ignores the bottom 25% and top 25%, making it ideal for skewed data or datasets with outliers. It's the basis for box plots and the 1.5×IQR outlier rule. **Mean absolute deviation** averages absolute deviations without squaring, giving a more intuitive "average distance from the mean." **Coefficient of variation** normalizes spread by the mean, enabling comparison across different measurement scales.
Bessel's correction (dividing by n−1 instead of n) corrects for the fact that sample variance tends to underestimate population variance. When you compute deviations from the sample mean (rather than the true population mean), the deviations are systematically too small because the sample mean is closer to the data than the population mean would be. Dividing by n−1 compensates for this bias exactly.
Skewness and kurtosis describe aspects of distribution shape that dispersion measures miss. A symmetric distribution with heavy tails (high kurtosis) looks very different from one with light tails, even if their standard deviations are identical. In finance, excess kurtosis quantifies "tail risk" — the probability of extreme events beyond what a normal distribution would predict. The 2008 financial crisis was partly a kurtosis problem: models assumed normal distributions when actual returns had much heavier tails.
Use IQR when your data is skewed or contains outliers. IQR is resistant to extreme values because it only considers the middle 50% of data. Standard deviation is more informative for symmetric distributions without outliers.
Sample variance divides by n−1 (Bessel's correction) to give an unbiased estimate of the population variance. Population variance divides by n and is used only when you have data for the entire population, not a sample.
CV expresses the standard deviation as a percentage of the mean, allowing comparison of variability across different scales. For example, a CV of 10% for both exam scores and heights means they have similar relative variability, even though their actual SDs differ dramatically.
Skewness measures asymmetry. Positive skewness (> 0.5) means a right tail (some very high values). Negative skewness (< −0.5) means a left tail. Values near 0 indicate approximate symmetry. Income data typically has strong positive skewness.
Any value below Q1 − 1.5×IQR (lower fence) or above Q3 + 1.5×IQR (upper fence) is flagged as an outlier. This is the same rule used in box plots. It's based on the middle 50% of data and is resistant to the outliers it's designed to detect.
The dispersion index (VMR = variance/mean) indicates whether data is over-dispersed (VMR > 1), equi-dispersed (VMR ≈ 1), or under-dispersed (VMR < 1) relative to a Poisson distribution. It's commonly used for count data.