How to Run ANOVA in Excel (Step by Step)

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-10-03

Running an ANOVA in Excel takes about two minutes once the Data Analysis ToolPak is enabled. You arrange each group in its own column, open Data > Data Analysis > Anova: Single Factor, and Excel returns a table with sums of squares, degrees of freedom, mean squares, the F statistic and a p-value. This guide walks through the whole process and shows how to read every cell of that output.

Quick Answer

Enable the ToolPak first: File > Options > Add-ins > Excel Add-ins > Go > check Analysis ToolPak [1].
Lay your data out in columns, one column per group, with a header row.
Go to Data > Data Analysis > Anova: Single Factor.
Set the Input Range to include headers, tick Labels in First Row, set Alpha (0.05 is standard), and pick an Output Range.
Read the F value and the P-value in the Between Groups row. If P-value < Alpha, the group means differ significantly.

Before You Start

You need three things in place before an ANOVA in Excel will run.

The Analysis ToolPak add-in. It ships with Excel but is not switched on by default. Go to File > Options > Add-ins, choose Excel Add-ins in the Manage dropdown, click Go, tick Analysis ToolPak, and click OK [1]. After that, a Data Analysis button appears on the Data tab. If you already see it, you are set.

Data in the right shape. A one-way ANOVA compares the means of three or more independent groups. Put each group in its own column, one observation per row, with a label in the top cell. Do not stack all the values in a single column with a separate group label column, because Anova: Single Factor expects the column layout.

A question with one factor. One-way means one categorical factor with several levels, for example three fertilizers, four teaching methods, or five suppliers. If you have two factors, you need a different tool.

If you want to understand what the output table is built from before you run it, the ANOVA table explained walkthrough covers the components and formulas.

Step by Step

Enter your data. Put group 1 in column A, group 2 in column B, group 3 in column C. Row 1 holds the group names. Every column should have the same number of observations for a balanced design, though Excel tolerates unequal group sizes.

Open the tool. Click the Data tab, then click Data Analysis in the Analysis group. In the list, select Anova: Single Factor and click OK [1].

Set the input range. In the Input Range box, type or select the full block including headers, for example $A$1:$C$9. Tick Labels in First Row so Excel treats row 1 as names, not data.

Choose the alpha level. Alpha is your significance threshold. Leave it at 0.05 unless your field uses a different convention.

Pick an output location. Select Output Range and enter a single cell such as $E$1. Excel writes the table starting there. You can also send it to a new worksheet.

Click OK. The ANOVA table appears immediately.

Read the result. Look at the Between Groups row. The F column holds the test statistic and the P-value column holds the probability of seeing an F that large if all population means were equal.

Worked Example

A grower tests three fertilizers on eight plants each and records growth in centimeters. The data sit in A1:C9, and the ANOVA output is written to E1.

	A	B	C	E	F	G	H	I	J
1	Fertilizer A	Fertilizer B	Fertilizer C	ANOVA Output
2	12.4	14.1	11.2	Source of Variation	SS	df	MS	F	P-value
3	13.1	15.3	12.8	Between Groups	`=DEVSQ(A2:C9)-F4` -> 42.05	2	`=F3/G3` -> 21.02	`=H3/H4` -> 52.11	`=F.DIST.RT(I3,G3,G4)` -> 7.20703E-09
4	11.8	14.8	10.9	Within Groups	`=DEVSQ(A2:A9)+DEVSQ(B2:B9)+DEVSQ(C2:C9)` -> 8.47	21	`=F4/G4` -> 0.40
5	12.9	15.9	12.1	Total	`=F3+F4` -> 50.52	`=G3+G4` -> 23
6	13.5	14.4	11.7
7	12.2	15.1	12.4
8	13.8	14.7	11.5
9	12.6	15.6	12.0

The formulas behind the table are worth knowing, because they let you rebuild the output by hand or check it.

Between-groups sum of squares measures how far each group mean sits from the grand mean, weighted by group size:

$$SS_{between} = \sum_{j=1}^{k} n_j (\bar{x}_j - \bar{x})^2$$

Within-groups sum of squares adds up the spread inside each group:

$$SS_{within} = \sum_{j=1}^{k} \sum_{i=1}^{n_j} (x_{ij} - \bar{x}_j)^2$$

The F statistic is the ratio of the two mean squares:

$$F = \frac{MS_{between}}{MS_{within}} = \frac{SS_{between} / (k-1)}{SS_{within} / (N-k)}$$

Here $k = 3$ groups and $N = 24$ observations, so $df_{between} = 2$ and $df_{within} = 21$.

Reading the numbers. The between-groups mean square is 21.02 and the within-groups mean square is 0.40. Their ratio gives F = 52.11. The p-value is 7.20703E-09, which is about 0.000000007. That is far below 0.05, so you reject the null hypothesis that all three fertilizers produce the same average growth. The data are consistent with at least one fertilizer differing from the others.

What the test does not tell you is which fertilizer differs. For that you need a post hoc comparison such as Tukey HSD, which is covered in the one-way ANOVA worked example.

Other Ways to Do It

Formulas only. You can build the whole table yourself with DEVSQ, F.DIST.RT and simple division, as the worked example shows. This is useful when you want the calculation visible on the sheet or need to check the ToolPak output.

The ANOVA Calculator. If you just want the numbers and not the spreadsheet, the ANOVA Calculator takes group data and returns the same F and p-value without any setup.

Manual means first. Before running the test, it helps to compute each group mean and see the spread. The mean guide and the variance guide cover those steps, and variance is the quantity the within-groups term is built from.

Chart the groups. A quick column chart of the three means makes the pattern obvious before you look at any p-value. See how to make a chart in Excel.

Troubleshooting

Data Analysis button is missing. The ToolPak is not enabled. Go back to File > Options > Add-ins > Excel Add-ins > Go and tick Analysis ToolPak [1].

"Input range contains non-numeric data" error. A text value or a blank cell is inside the range you selected. Check that every data cell holds a number and that Labels in First Row is ticked if row 1 contains names.

F is negative or the table looks scrambled. You probably selected the wrong input range or left Labels in First Row unticked, so Excel treated a header as a value.

Unequal group sizes. Anova: Single Factor handles them, but the df and MS values change. Do not assume the balanced formulas above still apply exactly.

P-value shows as 0. Excel displays very small probabilities in scientific notation or rounds them to 0 in a narrow column. Widen the column or format the cell as Scientific.

Common Mistakes

Running the test on two groups. A one-way ANOVA with two groups gives the same p-value as a two-sample t-test, but the t-test is the clearer choice. Use ANOVA for three or more groups.
Stacking data in one column. Anova: Single Factor expects one column per group. A single value column plus a label column will not work with this tool.
Forgetting Labels in First Row. If your range includes headers and the box is unticked, Excel reads the header text as data and either errors or produces nonsense.
Stopping at the p-value. A significant F says the means are not all equal. It does not say which pair differs. Run a post hoc test.
Ignoring the assumptions. ANOVA assumes independent observations, roughly normal residuals within groups, and similar variances across groups. Check group standard deviations before trusting a borderline result.
Re-running until it is significant. Dropping an outlier or a group after seeing the p-value inflates the false positive rate. Decide the analysis before you look at the result.

Limitations

Excel's Anova: Single Factor gives you the F test and nothing else. There is no built-in post hoc procedure, no effect size, no assumption diagnostics, and no Welch correction for unequal variances. You get the omnibus test and you build everything else yourself. That is fine for a quick check but thin for a formal report.

The tool also assumes independent samples. If your observations are paired, repeated on the same subjects, or nested in clusters, a one-way ANOVA on independent groups is the wrong model and the p-value will mislead you. The same applies when variances differ sharply across groups, where the standard F test can be too liberal or too conservative.

Frequently Asked Questions

How do I enable the Data Analysis ToolPak in Excel?

Go to File > Options > Add-ins. In the Manage dropdown at the bottom, select Excel Add-ins and click Go. Tick the box next to Analysis ToolPak and click OK [1]. A Data Analysis button then appears on the Data tab, and Anova: Single Factor is one of the tools listed inside it.

What is the difference between one-way and two-way ANOVA in Excel?

One-way ANOVA tests a single factor with several levels, such as three fertilizers. Two-way ANOVA tests two factors at once and can also test whether they interact. Excel's ToolPak offers Anova: Single Factor and Anova: Two-Factor With Replication and Anova: Two-Factor Without Replication as separate tools [1].

How do I interpret the p-value in the ANOVA output?

The P-value column in the Between Groups row is the probability of observing an F at least that large if all group means were truly equal. If it is below your alpha, usually 0.05, you reject that null hypothesis. In the fertilizer example the p-value is 7.20703E-09, so the means clearly differ.

Can I run an ANOVA in Excel without the ToolPak?

Yes. You can compute the sums of squares with DEVSQ, divide by the degrees of freedom to get mean squares, take the ratio for F, and use F.DIST.RT for the p-value. The worked example above shows each formula. It takes longer but leaves the calculation visible on the sheet.

Does a significant ANOVA tell me which groups differ?

No. A significant F only says the group means are not all equal. To find which specific pairs differ you need a post hoc test such as Tukey HSD, which Excel does not include in the ToolPak. You can compute it manually or use a dedicated statistics package.

References

Use the Analysis ToolPak to perform complex data analysis | Microsoft Support