SKEW.P in Google Sheets: Population Skewness Examples
Use SKEW.P in Google Sheets for population skewness, check third-moment arithmetic and distinguish its coefficient from adjusted sample SKEW.
Verification status: See the evidence and test scope described in this guide. How formula examples are checked
Quick answer: if A2:A6 contains 2, 2, 2, 2 and 7, =SKEW.P(A2:A6) has expected result 1.5. It describes positive asymmetry in this complete illustrative population of five values.
Syntax and optional additional inputs
=SKEW.P(value1, [value2, ...])
The Google SKEW.P reference displays a short two-position syntax, but its examples explicitly demonstrate both a single range and a range followed by two additional numbers. The syntax above makes that evidenced variable-length usage clear.
value1: first numeric value or range to include.- Additional arguments: more values or ranges forming the same dataset.
Google documents text as ignored and says the function supports an arbitrary number of arguments despite a nominal 30-argument limit. No omitted numerical default is involved: optional arguments add observations rather than change the calculation mode.
SKEW.P is for a dataset treated as the entire population. That is an analytical choice about what the data represent, not a statement that a large sample automatically becomes the whole population.
Worked example: population asymmetry
Enter this complete synthetic population in A2:A6:
| Cell | Value |
|---|---|
| A2 | 2 |
| A3 | 2 |
| A4 | 2 |
| A5 | 2 |
| A6 | 7 |
In C2, enter:
=SKEW.P(A2:A6)
Expected result: 1.5. Four values sit below the mean of 3 and one extends farther above it. Skewness weighs cubed deviations, so the one distant positive deviation dominates.
For a reflected dataset -2, -2, -2, -2, -7, the expected coefficient is -1.5. For 1, 2, 3, 4, 5, it is 0. These coefficients are dimensionless; formatting 1.5 as 150% would give it a probability interpretation it does not have.
Independent third-moment arithmetic
The original population’s deviations from its mean are -1, -1, -1, -1 and 4. Their squared sum is 20 and cubed sum is 60. Divide each by the population count 5 to obtain m2 = 4 and m3 = 12.
The coefficient is:
=(60/5)/(20/5)^1.5
Expected result: 1.5. This is the population third standardized central moment. NIST’s skewness discussion gives this coefficient separately from its adjusted sample counterpart.
As a documentation check, Google’s eight-value example 2, 5, 8, 13, 10, 18, 23, 26 gives 0.2763070768. Independent population-moment arithmetic reproduces that value. Adding 30 and 40 reproduces the second printed result, 0.4621754338.
Advanced application: inspect an expanded population
If a sixth valid member with value 12 is added to the original population, use:
=SKEW.P(A2:A6,12)
Expected result: 1.1222634355. The changed mean and moments produce a different coefficient. A larger maximum does not guarantee a larger skewness coefficient.
This mixed range-and-scalar pattern is useful when a new measurement has not yet been appended to the main range. Do not include the same observation both inside the range and as an additional scalar; that would count it twice and change its effective weight.
For comparison across units, adding 100 to every original value or multiplying every original value by 2 preserves the expected coefficient 1.5. Multiplying by a negative factor reverses its sign. These properties follow from how second and third moments transform.
Why SKEW produces a different number
For the original five values, adjusted sample SKEW has expected result 2.2360679775, rather than 1.5. Its adjustment depends on the observation count. This distinction is supported by comparing Google’s SKEW and SKEW.P numerical examples, not by assuming another product’s naming convention.
Use the same coefficient definition when comparing populations, studies or exported reports. Otherwise an apparent difference in asymmetry may partly reflect a different sample correction.
A zero coefficient does not guarantee symmetry or normality in every possible dataset. Positive and negative third-moment contributions can cancel. Inspect the underlying distribution as well as the summary, especially with extreme values or multiple clusters.
Errors and limitations
Google documents #DIV/0! for fewer than two supplied values and says text is ignored. A population with no variability makes the mathematical standardized-moment denominator zero; exact native constant-data handling was not established in this documentation-only check.
Blank handling, logical-value coercion and other native error cases remain unverified. Do not fill unknown measurements with zero unless zero is genuinely the observed value. Check that all included ranges refer to the intended population and do not overlap accidentally.
Availability and verification
SKEW.P is currently documented for rolling Google Sheets. Original introduction date, first release and last functioning numbered version were not established.
NOT_TESTED_IN_GOOGLE_SHEETS. Google’s individual function page and its numerical examples were inspected on October 17, 2026. Original and published coefficients were independently checked using central moments and numerical population-skewness calculations. No native execution or screenshot is claimed.