In AP Statistics, skewness describes the asymmetry of a quantitative distribution: a distribution is skewed right (positive) if its right tail is longer, skewed left (negative) if its left tail is longer, and symmetric if the two halves mirror each other.
Skewness is just a fancy word for which way a distribution leans. Picture a histogram. If the bars pile up on the left and a long tail trails off to the right, that's skewed right (also called positive skew). If the pile is on the right and the tail stretches left, that's skewed left (negative skew). If the left half is basically a mirror image of the right half, it's symmetric (no skew).
Under learning objective AP Stats 1.6.A, shape is one of the things you always describe about a quantitative distribution, alongside center, variability, and unusual features like outliers, gaps, and clusters. Here's the intuition that makes it click: the tail points toward the direction of the skew. A few unusually large values (think a handful of $175,000 salaries in a company full of $40,000 earners) drag the tail to the right, so the distribution is skewed right. Those extreme values also pull the mean toward the tail, which is why skewness is the reason mean and median split apart.
Skewness shows up in three different units, which is what makes it more than a vocab word. In Unit 1, AP Stats 1.6.A and 1.9.A/1.9.B ask you to describe and compare distributions, and shape (including skew) is a required part of any complete description. In Unit 7, skewness affects the normality condition for a two-sample t-test (AP Stats 7.8.C): if the data are skewed, you lean on large sample sizes (both n greater than 30) to trust that the sampling distribution is approximately normal. In Unit 9, the same kind of shape reasoning shows up when you check conditions for inference on a regression slope (AP Stats 9.4.C). So skewness is both a describing tool and a gatekeeper for whether your inference procedure is even valid.
Keep studying AP Statistics Unit 1
Outlier (Unit 1)
Outliers and skew are best friends. A cluster of extreme values on one side both creates a long tail and pulls the mean toward it, so spotting skew often means spotting outliers, and removing them usually makes a distribution more symmetric.
Box Plot (Unit 1)
A boxplot is a quick skew detector. When the right whisker is much longer than the left, or the median sits closer to Q1 than Q3, you're looking at right skew without ever drawing a histogram.
Normal Distribution (Units 1 & 7)
A Normal distribution is perfectly symmetric, so skewness is essentially the opposite of Normal. That matters for inference: when AP Stats 7.8.C asks if the sampling distribution is approximately normal, skewed data is exactly what forces you to check that n is at least 30.
Descriptive Statistics (Unit 1)
Skew decides which numbers honestly summarize your data. In a skewed distribution the mean gets dragged toward the tail, so median and IQR (which ignore the extremes) describe center and spread more fairly than mean and standard deviation.
On multiple choice, skewness usually hides inside a five-number summary or a boxplot. If you see Min=38,000, Median=48,000, Max=$175,000, that huge gap between Q3 and the max screams right skew, and the strongest answer is to use median and IQR instead of mean and standard deviation. Other stems give you quartiles and a far-off extreme and ask which unusual feature is present, or ask what removing values more than 3 SDs above the mean does to the shape. On free response, released FRQs lean on describing distributions from histograms (2019 Q1) and reasoning about unusually short or long values (2021 Q1). What you actually have to DO: name the shape using the right vocabulary (skewed left, skewed right, or symmetric), connect that shape to which summary statistics to report, and, in Units 7 and 9, use the shape to justify whether your normality condition holds.
Skewness is a shape of the whole distribution; an outlier is a single point that's unusually far from the rest. They're related (outliers often cause skew) but they're not the same thing. You can have a skewed distribution with no formal outliers, and a symmetric distribution can still have one extreme outlier on each side. On the exam, name the shape (skew) separately from any unusual features (outliers).
The tail tells you the direction: skewed right means the long tail points right, skewed left means it points left.
In a skewed distribution the mean gets pulled toward the tail, so the median is the more reliable measure of center.
Right skew usually means median and IQR are better summaries than mean and standard deviation.
A boxplot with one long whisker, or a median closer to one quartile than the other, is a fast way to spot skew.
Skew matters for inference: under AP Stats 7.8.C, skewed data requires both sample sizes greater than 30 for the normality condition to be met.
Always report shape (including skew), center, variability, and unusual features when you describe a quantitative distribution.
Skewness describes the asymmetry of a quantitative distribution. A distribution is skewed right if its right tail is longer, skewed left if its left tail is longer, and symmetric if its two halves mirror each other.
The mean gets pulled toward the tail; the median barely moves. So in a right-skewed distribution the mean is larger than the median, which is exactly why you report median and IQR for skewed data.
No. Skewness is the shape of the entire distribution, while an outlier is a single unusually large or small value. Outliers often cause skew, but you should name them as separate features when describing a distribution.
Not automatically. Under AP Stats 7.8.C, if the data are skewed you just need both sample sizes greater than 30 so the sampling distribution is still approximately normal. Skew only becomes a problem with small samples.
Compare the whiskers and the median's position. A much longer right whisker, or a median sitting closer to Q1 than Q3, signals right skew; a longer left whisker or a median near Q3 signals left skew.