A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Descriptive statistics is the practice of combining measures of center (mean, median, mode) and spread (range, IQR, standard deviation) with visual displays (histograms, boxplots) to characterize a distribution fully. A single summary statistic never tells the whole story — two datasets can share the same mean but differ drastically in spread, skewness, or outlier behavior. Effective description requires choosing the right statistics for the shape of the data: the median and IQR are more robust for skewed distributions, while the mean and standard deviation are most informative for roughly symmetric ones. This synthesis skill is the foundation for all further statistical inference: you must describe what you see before you can draw conclusions from it.
Give students several datasets with identical means but different shapes and spreads, and ask them to describe each fully. Practice writing narrative summaries that reference center, spread, shape, and outliers together. Pair every numerical summary with a graph so students learn that numbers and visuals complement each other.
You have already learned the individual tools — mean, median, and mode for center; range, IQR, and standard deviation for spread; boxplots and histograms for visual display. Descriptive statistics synthesis is the skill of combining these into a coherent account of a dataset. The key insight is that no single number tells the whole story. Two datasets can share exactly the same mean and still be completely different in character.
Consider two classrooms with the same average test score of 70. In Class A, every student scored between 65 and 75 — a small standard deviation and a tight boxplot. In Class B, half the students scored above 90 and half below 50 — a large standard deviation and long boxplot whiskers reaching both extremes. Reporting only the mean treats these classes as identical when they call for entirely different responses. This is why center and spread must always be reported together, and why the standard deviation or IQR are not optional extras.
Shape and outliers complete the picture. A distribution can be symmetric (mean ≈ median), right-skewed (long tail to the right, mean > median — a few unusually high values pull the mean up), or left-skewed (long tail to the left, mean < median). Outliers are data points that fall far from the bulk; they may be coding errors or genuine extremes worth investigating. The critical point is that you must look at a graph to see shape and outliers reliably. Numbers alone can hide them — two datasets can match on all five summary statistics and still have radically different distributions, as Anscombe's quartet famously demonstrates.
The practical rule for choosing which statistics to report follows directly from shape: for roughly symmetric distributions, report mean and standard deviation — they make full use of the data's numerical scale and are the foundation for later inferential methods. For skewed distributions or data with extreme outliers, report median and IQR — they describe the middle of the distribution without distortion from the tails. Reporting mean and standard deviation for heavily skewed salary data, for instance, gives a technically correct but practically misleading summary. Effective statistical description always matches the tool to the actual shape of the data you have.