How to Use the SKEW and SKEW.P Functions in Microsoft Excel to Measure Distribution Asymmetry

When summarising datasets in business spreadsheets, analysts typically report the mean (average) and standard deviation to describe the central tendency and spread of numbers. However, relying solely on these two metrics implicitly assumes that the data follows a symmetrical, bell-shaped normal distribution. In real-world business scenarios—such as customer household incomes, insurance claim payouts, website response times, and financial asset returns—distributions are frequently lopsided, with values clustered at one end and a long, tapering tail stretching toward extreme values at the other.

To quantify the degree and direction of asymmetry in a dataset, statisticians evaluate skewness. In Microsoft Excel, the SKEW and SKEW.P functions calculate this metric automatically. Measuring skewness is vital for financial risk modelling, determining whether to report the mean or median to stakeholders, and testing whether data meets the normality assumptions required for predictive algorithms. This comprehensive guide breaks down the mathematics of skewness, compares sample versus population formulas, demonstrates practical calculation workflows, and explains how to interpret positive, negative, and near-zero scores.

What Is Skewness and Why Does It Matter?

Skewness measures the degree of asymmetry of a probability distribution around its arithmetic mean. Unlike variance, which measures dispersion regardless of direction, skewness indicates which side of the distribution contains the longer or fatter tail:

• Symmetrical Distribution (Skewness ≈ 0): The data is evenly distributed around the mean, resembling a classic Gaussian bell curve. The mean, median, and mode coincide closely.
• Positively Skewed (Right-Tailed, Skewness > 0): The majority of observations cluster at lower values, but a handful of exceptionally large numbers stretch a long tail out to the right (such as executive executive compensation or lottery winnings). In a right-skewed distribution, the mean is pulled higher than the median.
• Negatively Skewed (Left-Tailed, Skewness < 0): Most observations cluster at high values, but occasional severe negative drops stretch a tail out to the left (such as retirement ages, lifespan longevity, or customer satisfaction scores on a high-performing product). In a left-skewed distribution, the mean is pulled lower than the median.

SKEW vs SKEW.P: Choosing the Correct Formula

Microsoft Excel offers two specific skewness functions depending on your data sample structure:

• SKEW (Sample Skewness): Evaluates a sample drawn from a broader population ($n ≥ 3$). It incorporates an adjustment factor ($n / ((n-1)(n-2))$) to produce an unbiased estimator of population skewness. This is the standard formula used in virtually all corporate analytics and academic research.
• SKEW.P (Population Skewness): Introduced in Excel 2013, this function calculates the true skewness when your dataset represents the entire complete population under study (such as all 50 US states, all 500 companies in the S&P 500, or every employee in your organization).

Function Syntax and Data Constraints

Both functions accept up to 255 numbers, cell ranges, or arrays:

=SKEW(number1, [number2], ...)

=SKEW.P(number1, [number2], ...)

Key Operational Constraints

• Minimum Data Points: SKEW requires at least three numeric observations ($n ≥ 3$) to evaluate degrees of freedom without division by zero. SKEW.P requires at least two observations ($n ≥ 2$).
• Non-Zero Variance: The dataset must possess non-zero standard deviation. If every cell in your range contains the identical value (e.g., all 5s), both functions return a #DIV/0! error.
• Text and Blanks: Empty cells, text labels, and logical values (TRUE/FALSE) inside referenced ranges are automatically ignored.

Step-by-Step Practical Example: Evaluating Household Customer Incomes

Suppose you are analysing customer annual household income data stored in cells B2:B25 to determine whether to present the mean or median to your marketing leadership team.

Step 1: Calculate the Sample Skewness

1. Click into an empty summary cell (e.g., D2).
2. Enter the sample skewness function:

   =SKEW(B2:B25)

3. Press Enter. Excel returns a numeric score (for instance, +1.85).

Step 2: Compare Mean and Median to Validate

1. In cell D3, calculate the average:

   =AVERAGE(B2:B25)

   (Returns £72,000).
2. In cell D4, calculate the median:

   =MEDIAN(B2:B25)

   (Returns £48,000).

Because the skewness value is substantially positive (+1.85 > 1.0), the high average is artificially inflated by a few wealthy outlier households. The analyst can confidently report that the median (£48,000) provides a more accurate representation of typical customer purchasing power.

Standard Rules of Thumb for Interpreting Skewness Values

Statisticians use standard benchmark rules to classify distribution asymmetry:

Skewness Range	Distribution Shape	Analytical Implication
Between -0.5 and +0.5	Fairly Symmetrical	Approximates normality. The mean is reliable.
Between -1.0 and -0.5 or +0.5 and +1.0	Moderately Skewed	Noticeable tail. Check both mean and median.
Less than -1.0 or Greater than +1.0	Highly Skewed	Heavy asymmetry. Report the median; consider log-transformation for regression.

Troubleshooting Common Errors with SKEW and SKEW.P

If your skewness calculation produces spreadsheet errors, review these common causes:

• #DIV/0! Error: Occurs when your dataset contains fewer than three data points in SKEW (or fewer than two in SKEW.P), or when all values are identical so the variance is zero. Ensure your range contains varying numbers.
• #VALUE! Error: Appears if you type non-numeric text strings directly into function arguments rather than referencing spreadsheet cell ranges.
• Zero Skewness Misinterpretation: A skewness score of zero indicates directional symmetry, but it does not guarantee a bell curve. Bimodal or uniform distributions can also be symmetrical with near-zero skewness; always generate a histogram chart to inspect distribution geometry visually.