Right-Skewed vs Left-Skewed Histograms
July 9, 2026
Right-Skewed vs Left-Skewed Histograms
People mix these two up constantly. The trick is to stop thinking about where the bars pile up and start thinking about where the tail points. The tail is the long, thin stretch of bars that trails off to one side. Whichever way the tail points is the name of the skew.
That single rule sorts out most of the confusion. Right-skewed means the tail points right. Left-skewed means the tail points left. Simple.
Right-skewed (positive skew)
In a right-skewed shape, the tall bars sit on the left and a long tail stretches out to the right. The peak is on the low end, and a handful of large values pull the graph out toward the high end.
Because those big values on the right tug the average upward, the mean lands higher than the median. The full order runs mode, then median, then mean, going left to right. The mode sits under the peak, the median is in the middle, and the mean gets dragged rightward by the tail.
Income is the classic case. Most people cluster in a normal range, but a few very high earners stretch the tail way out, so the average income sits above the typical income. House prices behave the same way. So do wait times, since most waits are short but a few drag on much longer than the rest.
You can see this instantly when you plot it. Drop your own numbers into the histogram maker and watch which side the tail leans toward. The dedicated right-skewed histogram guide breaks the shape down further.
Right skew, skewed to the right, positive skew: one shape, three names
People search for this shape a dozen different ways and worry the words mean different things. They do not. Right skew, skewed to the right, positive skew, and a right-skewed distribution all point to the exact same picture: a peak on the low end and a long tail dragging out to the right.
So when a question asks what it means for data to be skewed to the right, the answer is short. Most of the values sit low, a few large values stretch the high end, and the average gets pulled up above the middle value. A right-skewed distribution is not a rare or broken shape. It is one of the most common shapes real data takes, which is why income, prices, and time-based numbers keep showing up as examples.
The only trap is the name feeling backward. The bars are tall on the left, so people want to call it left-skewed. Ignore the bars. Read the tail. The thin tail points right, so the shape is skewed to the right, full stop.
Left-skewed (negative skew)
A left-skewed shape is the mirror image. The tall bars sit on the right, and the long tail stretches out to the left. The peak is on the high end this time, and a few small values pull the graph out toward the low end.
Here the small values on the left drag the average down, so the mean lands lower than the median. The order flips to mean, then median, then mode. The mean is on the left, pulled by the tail, and the mode sits under the peak on the right.
Scores on an easy test are a good example. Most students score high and bunch up near the top, while a few low scores create the tail on the left. Age at retirement works the same way, since most people retire around a common age but a few leave much earlier. Five-star ratings also skew left, because ratings pile up at the top and only a few low ones trail off. The left-skewed histogram guide covers this shape in detail.
Left skew, skewed to the left, negative skew: also one shape
The left side has the same pile of names, and they all mean one picture. Left skew, skewed to the left, negative skew, a negatively skewed histogram, and a left-skewed distribution are five labels for a peak on the high end with a tail dragging out to the low end.
"Negative" throws people because nothing about the data has to be negative. The word comes from the sign of the skewness number, which we get to below. A left-skewed set of test scores is all positive numbers and still counts as negative skew.
The backward-name trap works in reverse here too. The tall bars sit on the right, so people want to call it right-skewed. Read the tail, not the bars. The thin tail points left, so the shape is skewed to the left.
Graph, distribution, curve, or histogram: does the word matter?
Not for the shape. A right-skewed graph, a right-skewed distribution, a right-skewed curve, and a right-skewed histogram all describe the same lopsided pattern. The words differ only in what is being drawn.
A histogram is the specific bar chart of binned counts, the thing the histogram maker builds. A graph is the loose everyday word for any of these pictures. A distribution is the underlying pattern the data follows, whether or not anyone has drawn it. A curve is the smooth line you get if you traced the tops of the bars with infinitely many values.
So the skew rule carries across all of them without changing. Tail right means right-skewed, whatever you call the picture.
Putting a number on skew
Eyeballing the tail works fine most of the time. When you need to report skew rather than describe it, there is a simple formula, Pearson's second skewness coefficient:
skewness = 3 x (mean - median) / standard deviation
It formalizes the rule you already know. The mean chases the tail, so the further the mean sits from the median, the stronger the skew, and the sign tells you the direction.
Work it on sixteen wait times in minutes: 1, 1, 2, 2, 2, 3, 3, 3, 3, 4, 4, 5, 6, 8, 12, 19.
- mean = 4.88
- median = 3
- standard deviation = 4.70
3 x (4.88 - 3) / 4.70 = 1.20
Positive, so the data is right-skewed, which matches the two long waits at 12 and 19 minutes stretching the tail. Reading the result:
| Skewness value | What it means |
|---|---|
| Above +1 | Strong right skew |
| +0.5 to +1 | Moderate right skew |
| -0.5 to +0.5 | Roughly symmetric |
| -1 to -0.5 | Moderate left skew |
| Below -1 | Strong left skew |
A left-skewed set runs the formula the same way and comes out negative, because the mean falls below the median. That sign is the whole reason left skew is called negative skew.
Spotting skew in a box plot
Histograms are not the only chart that shows skew, and box plots often come up in the same coursework.
In a box plot, look at the median line inside the box and the two whiskers. If the median sits closer to the bottom of the box and the upper whisker is longer, the data is skewed right. Flip both and it is skewed left. A median centered in the box with even whiskers means roughly symmetric.
It is the same idea as the histogram, just compressed. The long whisker is the tail.
Side by side
Here is the whole comparison in one place.
| Feature | Right-skewed (positive) | Left-skewed (negative) |
|---|---|---|
| Tail points | Right | Left |
| Tall bars / peak | Left (low end) | Right (high end) |
| Mean vs median | Mean greater than median | Mean less than median |
| Order, low to high | Mode, median, mean | Mean, median, mode |
| Everyday examples | Income, house prices, wait times | Easy-test scores, age at retirement, five-star ratings |
The two rows that matter most are the tail direction and the mean-vs-median rule. If you remember that the tail names the skew, and that the mean always chases the tail, you have both shapes covered.
A quick way to remember it
The mean follows the tail. That is the whole thing. Wherever the long tail goes, the mean gets pulled in that direction, away from the median.
Right tail, mean goes right, so mean beats median. Left tail, mean goes left, so mean falls below median. The median stays calmer in the middle because it does not care about extreme values the way the mean does.
If you ever forget which is which, look at the tail and say its direction out loud. That is the skew name, no second-guessing.
Which average should you report?
Skew is not just a shape fact. It changes which number honestly describes your data.
When a histogram is skewed, the mean gets dragged toward the tail and stops representing a typical value. Report the mean on right-skewed income and it looks like everyone earns more than they do, because a few high earners pull it up. The median ignores those extremes and stays near the bulk of the data, so it is the fairer summary for either skew.
The rule of thumb: use the mean for roughly symmetric data, and switch to the median once a clear tail shows up. If you are not sure how symmetric your data is, plot it first. The histogram maker reports the mean and median together, and the gap between them tells you the story. A big gap means real skew, and it points in the direction of the tail. A small gap means the shape is close to symmetric and the mean is safe to use.
Common questions
What does right-skewed mean? It means the tail of the distribution stretches out to the right while most of the data piles up on the left. The name follows the tail, not the peak, so a shape that is tall on the left and thins out toward the right is right-skewed. It is also called positive skew.
What does it mean when data is skewed to the right? The same thing as right-skewed. Most values are small to moderate, a few large ones stretch the high end, and the mean sits above the median. Skewed to the right, right skew, and positive skew are three labels for one shape.
What is a right-skewed distribution? A distribution whose values cluster at the low end with a tail running toward the high values. Income and house prices are textbook right-skewed distributions: a large group near the middle and a small number of very high values pulling the average up.
Is a right-skewed histogram positive or negative skew? Positive. Right-skewed and positive skew are two names for the same shape, the one with the tail pointing right and the mean above the median.
How do you tell if data is skewed left or right? Find the long thin tail and read its direction. Tail to the right is right-skewed, tail to the left is left-skewed. The peak sits on the opposite side from the tail.
What does a right-skewed histogram tell you? That most values are low to moderate with a few unusually large ones stretching the top end. It is the fingerprint of income, prices, and wait times, where a small number of big values sit far above the crowd.
What is a negatively skewed histogram? The same thing as a left-skewed histogram: tall bars on the right, a tail running left, and a mean that lands below the median. It is called negative because Pearson's skewness formula returns a negative number for it, not because the data contains negative values.
What is the difference between a left-skewed and right-skewed graph? Only the direction the tail points, and everything that follows from it. A right-skewed graph peaks low and tails high with the mean above the median. A left-skewed graph peaks high and tails low with the mean below the median.
How do you calculate skewness? Pearson's second coefficient is the quick one: 3 x (mean - median) / standard deviation. Positive means right-skewed, negative means left-skewed, and anything between -0.5 and +0.5 is close enough to symmetric.
How do you tell if a box plot is skewed right? The median line sits low in the box and the upper whisker runs longer than the lower one. That long upper whisker is the same tail a histogram would show on the right.
Does skew change which average to use? Yes. The stronger the skew, the more the mean misleads, so lean on the median for skewed data and keep the mean for symmetric data.
Can a histogram be skewed and have two peaks? It can look lopsided and still hide a second peak. If your shape has two humps rather than one clean tail, read what two peaks in a histogram mean before you call it skewed, because a bimodal shape is a different story.
Try it on your own data
Reading about shapes only gets you so far. The fastest way to lock this in is to plot real numbers and see the tail form.
Paste a column of values into the histogram maker and the shape shows up right away. If you want the raw counts behind the bars, the frequency distribution table maker lays out each bin and how many values fall in it, which makes the skew easy to spot in the numbers themselves.
Skew is just one of the common patterns. Once you can read it, the other shapes get easier too. The shapes overview walks through the full set, including the bell-shaped and symmetric cases where the mean and median line up and no tail pulls anything off center.