The Boxplot Explained Simply

Full transcript

Word for word what is said in the video — the same text as the subtitles. Times are minutes:seconds into the video.

0:00A promise: all of today's numbers, one picture ▶

Last time ended with a promise: all of today's numbers, one small picture. Here are our 250 heights again, lined up shortest to tallest. By the end of this video, this whole line will fold into one small picture: the boxplot.

0:21The five-number summary ▶

Five numbers can describe this entire crowd. Three of them you already know: the first quartile at 164.6 centimetres, the median at 170, and the third quartile at 175.4 — the cuts from last time. The other two are simply the ends of the line: the shortest person and the tallest. Minimum, quartiles, median, maximum — statisticians call this the five-number summary, and it is all a boxplot needs.

0:54The fold: one line becomes one picture ▶

Now watch, because this is the whole trick. The middle half of the line folds up into a box: from the first quartile to the third, with a line at the median. The box holds the middle fifty percent of everyone. From each end of the box, a whisker reaches out to the farthest person who still counts as ordinary. And whoever sits beyond that stays behind as a loose dot. One line of 250 people has become five strokes of ink.

1:29The fences: where ordinary ends ▶

So where exactly does ordinary end? Take the width of the box — the interquartile range, 10.8 centimetres — and step one and a half box- widths out from each end. Those fences land at 148.4 and 191.6 centimetres. Whiskers stop at the last person inside the fences, so their length is earned by the data — they are not a fixed percentage. In our crowd, two people sit just past the fences. That is a flag, not a verdict: these are real heights, simply worth a second look. Now bring back the two- metre man from video two. He lands far beyond the fence, flagged again. The z-score needed a bell curve to catch him — the boxplot just needed the fences.

2:25Reading the box: lopsided data at a glance ▶

Once you can draw it, you can read it. Remember the lopsided shapes from last time — waiting times, incomes? On a boxplot you spot that instantly: the median slides off-centre inside the box, and one whisker runs long. A lopsided box means lopsided data — no curve required. Why so much data leans like that is a story of its own, and it is coming in two videos.

2:56The superpower: five groups, one glance ▶

And here is the boxplot's superpower: comparison. Put five groups side by side as histograms, and you get chaos — five overlapping mountains. Put them side by side as boxplots, and the story reads itself: the medians step upward from group to group, and one group is plainly twice as spread out as its neighbours. Five distributions in one glance — that is why scientific papers are full of these.

3:27One box can hide two hills ▶

One honest warning before you trust any box. This dataset contains two separate groups — two hills in the data. But fold it into a boxplot, and the box looks perfectly ordinary: one median, one tidy box, nothing to see. The box cannot show two hills, because five numbers cannot draw two bumps. So before you trust any box, plot the raw points once.

3:58In one picture ▶

So, in one picture: five numbers — minimum, first quartile, median, third quartile, maximum. The box holds the middle fifty percent, with the median inside it. Whiskers reach the last data inside the fences, at one and a half interquartile ranges. Dots beyond the fences are flags, not verdicts. And a box can hide two hills, so plot the points once.

4:30What's next: mean vs median ▶

But notice what the box never showed you: the mean. The box is built on the median — so which of the two should you trust? Next time: mean versus median, and which one to use when. Tell me below — what is the strangest outlier your data ever flagged? And if you want to keep learning, subscribe — the next video will find you. Until then: happy learning.