Standard error tells you how much your sample average might differ from the true population average

Standard error measures the spread of sample means — it shows how much variation you'd expect if you took the same measurement many times. In practical terms, it's the margin of uncertainty around an average you've calculated from a subset of data. A smaller standard error means your average is more reliable; a larger one means more caution is warranted.

Excel does not have a single "standard error" button. Instead, you calculate it by dividing the standard deviation of your data by the square root of how many data points you have. The formula is: standard deviation ÷ √(sample size). Excel's built-in functions make this straightforward once you know which ones to use.

Key Takeaways

  • Standard error uses two pieces: the standard deviation of your data (found with STDEV.S) and the count of your data points (found with COUNT).
  • The complete formula in Excel is =STDEV.S(range)/SQRT(COUNT(range)), where range is the cells holding your numbers.
  • STDEV.S is for sample data (a subset of a larger population); use STDEV.P only if your data represents an entire population.
  • Standard error shrinks as your sample size grows, so larger datasets produce more confident averages.

The formula: standard deviation divided by the square root of sample size

Open your spreadsheet and locate the column or range containing your data. Let's say your numbers are in cells A2 through A51 (50 data points). Click on an empty cell where you want the result to appear — perhaps C2.

Type this formula: =STDEV.S(A2:A51)/SQRT(COUNT(A2:A51))

Press Enter. Excel calculates the standard deviation of your data, counts how many numbers are in that range, takes the square root of the count, and divides the first by the second. The result is your standard error.

If your data is scattered across non-adjacent cells, you can list them separately: =STDEV.S(A2:A10,C2:C15)/SQRT(COUNT(A2:A10,C2:C15)). Excel will include all of them in both the standard deviation and the count.

Why STDEV.S, not STDEV.P

Excel offers two standard deviation functions: STDEV.S and STDEV.P. The difference matters for standard error. Use STDEV.S when your data is a sample — a subset drawn from a larger group you're trying to understand. Use STDEV.P only when your data represents the entire population you care about.

In most real situations, you're working with a sample. If you measured 50 customers' satisfaction scores, those 50 are a sample of all possible customers. If you measured every single customer who walked into your store on one day, that's still a sample of all customers across all days. STDEV.P is rare; it applies when you have literally every data point that exists for your question.

Using STDEV.P when you should use STDEV.S will make your standard error artificially small and give you false confidence in your average.

Breaking down each part of the formula

STDEV.S(A2:A51) calculates how spread out your numbers are. It measures the typical distance of each number from the average. A tight cluster of similar numbers produces a small standard deviation; scattered numbers produce a large one.

COUNT(A2:A51) counts how many numbers are in your range. It ignores empty cells and text. If you have 50 numbers, COUNT returns 50.

SQRT(COUNT(A2:A51)) takes the square root of that count. For 50 data points, SQRT returns about 7.07. This is why larger samples produce smaller standard errors — the denominator grows, making the final result smaller.

Dividing the standard deviation by this square root gives you the standard error. The formula works because standard error depends on both how variable your data is and how many observations you have to estimate the true average.

A worked example with real numbers

Suppose you measured the weight of 25 apples from a shipment. Your weights (in ounces) are in cells B2 through B26. The average weight is 6.2 ounces, but you want to know how confident you should be in that number.

In cell D2, type: =STDEV.S(B2:B26)/SQRT(COUNT(B2:B26))

Excel calculates: the standard deviation of your 25 weights is 0.8 ounces. The square root of 25 is 5. So the standard error is 0.8 ÷ 5 = 0.16 ounces. This means the true average weight of all apples in that shipment is likely within about 0.16 ounces of your measured 6.2 ounces — probably between 6.04 and 6.36 ounces. If you had measured 100 apples instead, the standard error would be smaller, and your estimate would be tighter.

Storing the formula for repeated use

If you calculate standard error often, create a helper row or column. Put your data range in one cell (as text, for reference), the standard deviation in another, the count in another, and the standard error in a fourth. This makes it straightforward to see each step and to reuse the structure for new datasets.

Alternatively, you can create a named formula. Go to the Formulas tab, select Define Name, and create a formula called StandardError that references a specific range. Then you can type =StandardError in any cell and get the result. This works best if your data always occupies the same cells.

Common mistakes and how to avoid them

The most frequent error is using STDEV.P instead of STDEV.S. This makes your standard error too small. Double-check which function matches your data: sample or population.

Another mistake is including text or empty cells in your range without realizing it. COUNT ignores both, so your sample size might be smaller than you think. If you have 50 cells but 5 are empty, COUNT returns 45, and your standard error will be slightly larger than it should be. Check your range carefully, or use a formula like =COUNTA(A2:A51)-COUNTBLANK(A2:A51) to verify the actual count of numbers.

A third mistake is forgetting the SQRT function. The formula must divide by the square root of the count, not the count itself. Omitting SQRT will make your standard error far too small.

When standard error matters in your spreadsheet

Standard error is useful when you're presenting an average and want to show how much uncertainty surrounds it. If you're reporting survey results, test scores, or measurements from a sample, standard error tells your audience how much the true average might differ from what you observed.

You can use standard error to build confidence intervals — ranges that likely contain the true average. A rough rule is that the true average is probably within two standard errors of your measured average, about 95 percent of the time. So if your average is 100 and your standard error is 5, you might say the true average is probably between 90 and 110.

Standard error is different from standard deviation. Standard deviation describes the spread of your actual data. Standard error describes the uncertainty in your average. Both are useful, but for different reasons.

Frequently Asked Questions

What's the difference between standard error and standard deviation?

Standard deviation measures how spread out your actual data points are. Standard error measures how much your calculated average might differ from the true average. Standard error is always smaller than standard deviation because it accounts for the fact that averaging reduces noise.

Do I use STDEV.S or STDEV.P?

Use STDEV.S for sample data — numbers drawn from a larger group. Use STDEV.P only if your data represents every single observation that exists for your question. Most real work uses STDEV.S.

Why does standard error get smaller with more data?

The formula divides by the square root of sample size. As you measure more items, that denominator grows, making the standard error smaller. More data means more confidence in your average.

Can I calculate standard error for text or categorical data?

No. Standard error applies only to numeric data — measurements, counts, or ratings on a scale. For categories like "yes/no" or "red/blue/green," use proportions or other methods instead.

What if my data has outliers that skew the standard deviation?

The standard error formula uses standard deviation, so outliers affect it. If you suspect outliers are distorting your results, check your data for entry errors first. If the outliers are real, you may want to report both the standard error and a note about the outliers, so readers understand the full picture.