How to Calculate Skewness in Excel

Key Insights

• Excel provides two built-in functions for skewness: SKEW() for sample data and SKEW.P() for population data—using the wrong one introduces bias into your analysis.
• Skewness values beyond ±1 indicate significant asymmetry that may require data transformation before applying parametric statistical methods.
• Manual formula calculation serves as validation and helps you understand exactly what Excel computes, which matters when defending your analysis to stakeholders.

Introduction to Skewness

Skewness measures the asymmetry of a probability distribution around its mean. In practical terms, it tells you whether your data leans left, leans right, or sits symmetrically balanced.

A positive skewness (right-skewed) means the tail extends toward higher values—think income distributions where most people earn moderate amounts but a few earn millions. A negative skewness (left-skewed) means the tail extends toward lower values—like exam scores where most students perform well but a few bomb completely. Zero skewness indicates perfect symmetry, though you’ll rarely see exactly zero in real data.

Why does this matter? Many statistical methods assume normally distributed data. When your data is significantly skewed, means become misleading, confidence intervals lose accuracy, and regression assumptions break down. Knowing your skewness value helps you decide whether to proceed with standard analysis or apply transformations first.

Understanding Excel’s Built-in SKEW Functions

Excel offers two skewness functions that look similar but serve different purposes:

SKEW() calculates sample skewness using the adjusted Fisher-Pearson standardized moment coefficient. Use this when your data represents a sample from a larger population—which covers 95% of real-world scenarios.

SKEW.P() calculates population skewness. Use this only when you have data for an entire population, not a sample. This is rare outside of census data or complete transaction logs.

The mathematical difference lies in the adjustment factor. Sample skewness includes a correction for bias that population skewness doesn’t need.

=SKEW(A1:A100)

This formula returns the sample skewness for values in cells A1 through A100. The function ignores text, logical values, and empty cells automatically.

=SKEW.P(A1:A100)

Same range, but calculates population skewness. The result will be slightly different—typically closer to zero than the sample version.

When to use which: Default to SKEW(). Only use SKEW.P() when you’re absolutely certain you have complete population data. If you’re analyzing survey responses, sales figures, or experimental measurements, you’re working with samples.

Step-by-Step Calculation Using SKEW()

Let’s work through a concrete example. Suppose you have monthly sales figures for a retail store in column B, rows 2 through 25 (24 months of data).

First, create a named range for cleaner formulas:

1. Select cells B2:B25
2. Click the Name Box (left of the formula bar)
3. Type “SalesData” and press Enter

Now calculate skewness:

=SKEW(SalesData)

Assume this returns 0.847. What does that mean?

The positive value indicates right skewness—your sales distribution has a tail extending toward higher values. Most months cluster around a typical sales figure, but occasionally you hit exceptional months that pull the average up.

A value of 0.847 falls in the “moderately skewed” range. It’s noticeable but not extreme. You could proceed with most analyses, though you should acknowledge the asymmetry in your reporting.

For comparison, calculate the mean and median:

=AVERAGE(SalesData)
=MEDIAN(SalesData)

With positive skewness, the mean will exceed the median. This confirms your skewness calculation—the high-value tail pulls the average upward while the median stays anchored at the 50th percentile.

Manual Skewness Calculation with Formulas

Understanding the underlying formula helps you validate Excel’s output and explain your methodology. The sample skewness formula is:

$$g_1 = \frac{n}{(n-1)(n-2)} \sum_{i=1}^{n} \left(\frac{x_i - \bar{x}}{s}\right)^3$$

Breaking this into Excel components:

=COUNT(SalesData)*SUM(((SalesData-AVERAGE(SalesData))/STDEV.S(SalesData))^3)/((COUNT(SalesData)-1)*(COUNT(SalesData)-2))

This is an array formula in older Excel versions. In Excel 365 or Excel 2021, it calculates automatically. In earlier versions, you’d need to press Ctrl+Shift+Enter instead of just Enter.

Let’s decompose this formula:

Step 1: Count the observations

=COUNT(SalesData)

Returns n, the number of data points.

Step 2: Calculate standardized values

=(SalesData-AVERAGE(SalesData))/STDEV.S(SalesData)

This creates z-scores for each value—how many standard deviations each point sits from the mean.

Step 3: Cube the standardized values

=((SalesData-AVERAGE(SalesData))/STDEV.S(SalesData))^3

Cubing preserves the sign (positive or negative) while amplifying extreme values.

Step 4: Sum and apply the adjustment factor The full formula sums these cubed values and multiplies by n/((n-1)(n-2)) to correct for sample bias.

Your manual calculation should match =SKEW(SalesData) exactly. If it doesn’t, check for hidden characters, filtered rows, or mixed data types in your range.

Visualizing Skewness with Histograms

Numbers tell part of the story. Histograms show the shape directly.

Using the Data Analysis ToolPak:

1. Enable the ToolPak: File → Options → Add-ins → Manage Excel Add-ins → Check “Analysis ToolPak”
2. Go to Data → Data Analysis → Histogram
3. Configure the dialog:
   • Input Range: Select your data (B2:B25)
   • Bin Range: Leave blank for automatic bins, or specify custom boundaries
   • Output Range: Choose where results appear
   • Check “Chart Output”

The resulting histogram reveals distribution shape visually. A right-skewed distribution shows bars clustered on the left with a tail stretching right. Left-skewed shows the opposite pattern.

Quick histogram alternative (Excel 2016+):

1. Select your data
2. Insert → Charts → Histogram

This creates an instant histogram without the ToolPak. Right-click the horizontal axis and select “Format Axis” to adjust bin width and overflow bins.

Add a vertical line at the mean and another at the median. When these lines diverge noticeably, you have visual confirmation of skewness. The direction of divergence matches the skewness sign.

Interpreting Results and Common Thresholds

Skewness values require context. Here are practical interpretation guidelines:

Implications for statistical analysis:

When skewness exceeds ±1, the mean becomes a poor measure of central tendency. Report the median instead, or report both with an explanation.

Regression analysis assumes normally distributed residuals. Highly skewed dependent variables often produce skewed residuals, violating this assumption. Check residual plots after fitting models.

Common transformations for positive skewness:

=LN(SalesData)           ' Natural log transformation
=SQRT(SalesData)         ' Square root transformation  
=SalesData^(1/3)         ' Cube root transformation

For negative skewness, reflect the data first:

=LN(MAX(SalesData)+1-SalesData)

After transformation, recalculate skewness. The goal is reducing the absolute value toward zero, not achieving exactly zero.

Practical Applications and Troubleshooting

Financial returns analysis:

Stock returns typically show slight negative skewness—small gains are common, large losses are rare but impactful. Calculate skewness on daily returns to quantify tail risk:

=SKEW(DailyReturns)

A skewness of -0.8 signals meaningful downside tail risk that variance alone doesn’t capture.

Quality control:

Manufacturing defect rates often show positive skewness. Most batches have few defects, but occasional batches fail badly. Skewness helps set realistic control limits:

=SKEW(DefectRates)

High positive skewness suggests using percentile-based limits rather than mean ± standard deviation.

Error handling:

The SKEW() function returns #DIV/0! when given fewer than three data points. Wrap calculations in error handling:

=IFERROR(SKEW(DataRange), "Insufficient data")

For more informative error messages:

=IF(COUNT(DataRange)<3, "Need 3+ values", SKEW(DataRange))

Dealing with outliers:

Extreme outliers inflate skewness dramatically. Before accepting a skewness value, check for data entry errors:

=MAX(DataRange)
=MIN(DataRange)

If these seem unreasonable, investigate before calculating skewness. One typo—entering 10000 instead of 100—can shift skewness from 0.3 to 3.0.

Consider calculating skewness with and without suspected outliers:

=SKEW(IF((DataRange>PERCENTILE(DataRange,0.01))*(DataRange<PERCENTILE(DataRange,0.99)),DataRange))

This excludes the top and bottom 1% of values. If skewness changes dramatically, your outliers are driving the result.

Skewness is one diagnostic among many. Combine it with kurtosis, visual inspection, and domain knowledge to make sound analytical decisions. The number itself means nothing without context—but with context, it guides you toward appropriate statistical methods and honest data reporting.