There's a recipe for describing any dataset, and it's only three steps. Centre, spread, shape — in that order — then put them in a sentence.
Step 1 — Centre: What's a Typical Value?
Quote the mean or the median — whichever is fairer. If the data is bunched up evenly, they'll sit close together and either is fine. If there's a long tail or an outlier dragging things, the median is the honest "typical". For 3, 4, 4, 5, 5, 5, 6 the mean and median both land near 5 — so "a typical value is about 5" is fair either way.
Step 2 — Spread: How Varied Is It?
Reach for the range — biggest minus smallest. A small range means everyone's similar (consistent); a big range means the values are all over the place. For that same list, 6 − 3 = 3, a fairly tight spread. Saying "the range is 3, so the results are pretty consistent" already tells the reader something the average can't.
Step 3 — Shape: How Do the Dots Sit?
Glance at the dot plot. Are they clustered in one lump, spread evenly along the line, or skewed — most values down one end with a tail trailing off the other way? Skew is the big tell: when the data has a tail, the mean gets tugged toward it, which is exactly why you'd quote the median instead.
Worked Example — Putting It Together
Suppose the data is 2, 2, 3, 3, 3, 4, 9. The mean is about 3.7 but the median is 3 — they disagree, a clue something's pulling the mean. The range is 9 − 2 = 7, and the dot plot shows a cluster low down with one lonely 9 off to the right: skewed. So a fair conclusion reads: "Most values sit around 3 (the median), the data is fairly spread (range 7) and it's skewed by one high value, so the median describes it better than the mean." That single sentence covers centre, spread and shape — and it's honest.