This page shows you how to run the common statistical tests in Excel, using both the Analysis ToolPak and worksheet functions, which output cell actually matters, and how to write the result in APA. When the deadline is tight, the data will not behave, or Excel is the wrong tool for the design, a statistician can run and interpret it on your own file.
Excel runs statistics in two ways, and knowing which to use is half the assignment. A worksheet function such as T.TEST or CORREL returns a single number in a cell. The Analysis ToolPak, a free add-in that ships with Excel, produces a full output table that looks like the report your course expects. Pick the wrong test for the design, lay the data out the wrong way, or read the wrong cell of the output, and the analysis is wrong no matter how tidy the sheet looks. This guide is part of our statistics homework help. It gives the exact menu path and function for each common test, points at the one value in the output that decides the result, and shows how to report it.
One honest limit up front. Excel is good for t-tests, one-way ANOVA, correlation, and chi-square, which covers most introductory courses. It is limited for advanced work: there is no built-in post-hoc test for ANOVA, no Levene's test for equal variances, and no direct significance test for a correlation coefficient. Where your design needs those, SPSS or R is the better tool, and this page says so at each test where it matters.
Match your research question and variable types to the right test, then jump to its guide.
| Your question | Your data | Test |
|---|---|---|
| Do two unrelated groups differ on average? | Continuous outcome, two independent groups | Independent-samples t-test |
| Did the same people change between two points? | Continuous outcome, two paired measures | Paired-samples t-test |
| Do three or more unrelated groups differ? | Continuous outcome, three or more groups | One-way ANOVA (single factor) |
| Are two categories associated? | Two categorical variables | Chi-square test of independence |
| Do two continuous variables move together? | Two continuous variables | Pearson correlation |
Repeated measures across three or more points, a post-hoc comparison, or a non-parametric test? Excel does not do those cleanly, so see the fallback table and consider SPSS or R.
Most statistical tests on this page use the Analysis ToolPak, a free add-in that comes with Excel but is switched off by default. You enable it once and it stays on.
When you click Data > Data Analysis a list appears with the tools you need, including t-Test in three versions, Anova: Single Factor, and Correlation. If the Data Analysis button is missing after you tick the box, close and reopen Excel.
How to lay out the data. Excel's statistical tools expect tidy columns. For a two-group comparison, put each group's scores in its own column, for example group A in column A and group B in column B, with a short label in the top cell of each. For one-way ANOVA, use one column per group. For correlation and the paired t-test, use two side-by-side columns of equal length, one row per case. Keep a header row and point the tool at it with the Labels in first row box ticked, which keeps the output readable. Excel cannot arrange the data for you, so a stacked single-column layout, which SPSS prefers, has to be split into columns first.
Set your alpha level in the tool's Alpha box, which defaults to 0.05, and choose an output location so the results table does not overwrite your data. The worksheet functions, T.TEST, CHISQ.TEST, CORREL, and PEARSON, need no add-in and return a single number, which is handy for a quick p-value or r but gives you no table to hand in.
The menu path, the function, the assumption checks, the exact cell to read, and how to report it. Example values follow standard teaching datasets.
Compares the means of one continuous outcome across two unrelated groups, for example test scores for a taught group and a control group.
Assumptions. A continuous outcome, two independent groups, no serious outliers, and approximate normality within each group. Excel has no Levene's test for equal variances, so you cannot check that assumption formally the way SPSS does. The safe default is the Unequal Variances version (Welch's t-test), which does not assume equal spread. If normality is badly broken, Excel has no built-in Mann-Whitney U test, so that is a point to move to SPSS or R.
What to read. In the output table, read t Stat, the df, and P(T<=t) two-tail. Report the two-tail value, not the one-tail value, unless your hypothesis is directional. Ignore t Critical, which is only the cut-off.
| Variable 1 | Variable 2 | |
|---|---|---|
| Mean | 78.40 | 72.10 |
| Variance | 52.36 | 61.88 |
| Observations | 20 | 20 |
| Hypothesized Mean Difference | 0 | |
| df | 38 | |
| t Stat | 2.428 | |
| P(T<=t) one-tail | 0.010 | |
| t Critical one-tail | 1.686 | |
| P(T<=t) two-tail | 0.020 | |
| t Critical two-tail | 2.024 |
Common mistakes
Outliers, a non-normal outcome, or a design Excel cannot test on your own dataset? Our statisticians will run the correct version, in Excel or SPSS as the data demands, and interpret it. Get a quote →
Compares two measurements taken on the same people, for example a score before and after training.
Assumptions. A continuous outcome measured twice on the same cases, matched row by row, no serious outliers, and normality of the difference scores, not the raw columns. Compute the differences in a helper column with =B2-A2 and eyeball them for skew and outliers. If the differences are badly non-normal, Excel has no Wilcoxon signed-rank test, so that is a point to move to SPSS or R.
What to read. Read t Stat, df, which equals the number of pairs minus one, and P(T<=t) two-tail. Do not read the Pearson Correlation row at the top of the table, which only describes how the two columns move together.
| Before | After | |
|---|---|---|
| Mean | 2.48 | 2.52 |
| Variance | 0.026 | 0.026 |
| Observations | 20 | 20 |
| Pearson Correlation | 0.972 | |
| Hypothesized Mean Difference | 0 | |
| df | 19 | |
| t Stat | 4.773 | |
| P(T<=t) one-tail | 0.0001 | |
| t Critical one-tail | 1.729 | |
| P(T<=t) two-tail | 0.0001 | |
| t Critical two-tail | 2.093 |
Common mistakes
Compares the means of one continuous outcome across three or more unrelated groups.
Assumptions. A continuous outcome, three or more independent groups, no serious outliers, normality within groups, and equal variances. Two honest Excel limits apply here. First, there is no post-hoc test: a significant ANOVA tells you the groups differ somewhere, but Excel will not tell you which pairs differ, so you cannot run Tukey or Games-Howell. Second, there is no Levene's test for equal variances. If you need post-hoc comparisons, which most write-ups require, run the ANOVA in SPSS or R instead.
What to read. In the lower ANOVA table read the Between Groups row for F and P-value, with the two df values, between-groups df then within-groups df. The SUMMARY table above it gives each group's mean and variance for the write-up.
| Source of Variation | SS | df | MS | F | P-value | F crit |
|---|---|---|---|---|---|---|
| Between Groups | 85.5 | 2 | 42.75 | 4.467 | 0.021 | 3.354 |
| Within Groups | 258.3 | 27 | 9.57 | |||
| Total | 343.8 | 29 |
Common mistakes
ANOVA with post-hoc comparisons, or an assumption that will not hold, is where Excel runs out of road. Send your file and we run it end to end and write the result. Get a quote →
Tests whether two categorical variables are associated, for example gender and a preferred learning format.
Assumptions. Both variables categorical, independent observations, and expected counts large enough, with no more than 20 percent of cells below an expected count of five. Excel does not warn you if that rule is broken, so check the expected table yourself. If a two-by-two table has small expected counts, Excel has no Fisher's Exact Test, so move to SPSS or R.
What to read. CHISQ.TEST returns only the p-value. To report the chi-square statistic and degrees of freedom you compute them yourself: df is (rows minus one) times (columns minus one), and the statistic is the sum of (observed minus expected) squared divided by expected across all cells, or =CHISQ.INV.RT(p, df).
| Cell | Formula | Result |
|---|---|---|
| Chi-square p-value | =CHISQ.TEST(B2:C3, B7:C8) | 0.485 |
| Degrees of freedom | =(2-1)*(2-1) | 1 |
| Chi-square statistic | =CHISQ.INV.RT(0.485, 1) | 0.487 |
Common mistakes
Measures the strength and direction of the linear relationship between two continuous variables.
Assumptions. Two continuous variables, a linear relationship, and no serious outliers, all of which you should check on a scatterplot first because Pearson only captures straight-line association. Here is the key Excel limit: CORREL and PEARSON return r but no significance test. Excel gives you the coefficient, not the p-value. For ordinal data or a monotonic but non-linear relationship you would use Spearman, which Excel has no direct function for.
What to read. Read the off-diagonal cell of the Correlation matrix, which is r. The diagonal is always one, because a variable correlates perfectly with itself. To get a p-value in Excel you either run Regression from the ToolPak, whose output reports the significance of the slope, or compute the t-statistic by hand as r * SQRT((n-2) / (1 - r^2)) and convert it with T.DIST.2T.
| Height | Jump distance | |
|---|---|---|
| Height | 1 | |
| Jump distance | 0.706 | 1 |
Common mistakes
Most statistics marks are lost not on the software but on ignoring a broken assumption. Excel checks fewer of them than SPSS does, so some rows below point straight to another tool. Check the assumption, and when it fails, switch to the matched approach rather than reporting an invalid result.
| Assumption or need | How Excel handles it | If it fails or is missing |
|---|---|---|
| Normality | No formal test; inspect a histogram of the values or differences | Excel has no non-parametric test, so use SPSS or R (Mann-Whitney, Wilcoxon, Kruskal-Wallis) |
| Equal variances | No Levene's test; use the Unequal Variances t-test as the safe default | Prefer the Unequal Variances (Welch) version, or check formally in SPSS |
| Post-hoc after ANOVA | Not available in Excel at all | Run the ANOVA in SPSS or R for Tukey or Games-Howell |
| Expected cell counts (chi-square) | You compute the expected table yourself; no warning is given | For small counts use Fisher's Exact Test in SPSS or R |
| Significance of a correlation | CORREL returns r only, no p-value | Use ToolPak Regression, or compute the t-statistic by hand |
| Repeated measures over 3+ points | No clean repeated-measures ANOVA | Use SPSS General Linear Model or R |
Procedures on this page were checked against the Microsoft Support documentation for the Analysis ToolPak and the T.TEST function.
When the data will not behave, an assumption fails, or Excel is the wrong tool for the design, a specialist runs it on your file and explains the output so you can defend it.
Upload your data file, the assignment brief and your deadline, then get a free quote from support.
Approve the price and a statistician who works in Excel every day starts right away.
Get the workbook, output, interpretation and write-up in your account, with free revisions if you need them.
Your work goes to a specialist who passed a subject-specific statistics test and a background check, and who works in Excel every day. What we deliver is a model answer and study aid for reference, and many students use it to learn the method and check their own analysis.
Everything stays private. Your data and details are confidential, we never contact your school, payments are secure, and the work is original and plagiarism-checked.
Get Excel help you can trust →"This company is wonderful. I was able to depend on them all semester for my Statistics assignments. Dependable with competitive prices. Would definitely use again!"
"I was a little skeptical at first but this is definitely the best decision I could have made when it comes to getting help with my homework. This is a legit business and the staff are very helpful and professional."
You have two routes. The function T.TEST(array1, array2, tails, type) returns the p-value: use type 1 for paired, type 2 for two independent samples with equal variances, and type 3 for two independent samples with unequal variances, with tails set to 2 for a standard two-tailed test. For the full output table, enable the Analysis ToolPak and choose Data, Data Analysis, then the matching t-Test tool, which reports t Stat and P(T<=t) two-tail.
On Windows go to File, Options, Add-ins. In the Manage box at the bottom choose Excel Add-ins and click Go, tick Analysis ToolPak, then OK. On a Mac go to Tools, Excel Add-ins, tick Analysis ToolPak, then OK. A Data Analysis button then appears on the right of the Data tab.
It is the two-tailed p-value for the t-test. If it is below your alpha level, usually 0.05, the difference between the two means is statistically significant. It is the value you report for an ordinary non-directional hypothesis. The one-tail value is only for a directional prediction, and t Critical two-tail is the cut-off, not the p-value.
Excel handles t-tests, one-way ANOVA, correlation and chi-square well, which covers most introductory courses. It is limited for advanced work: no built-in post-hoc test for ANOVA, no Levene's test, and no direct significance test for a correlation. For repeated-measures designs, post-hoc comparisons or non-parametric tests, SPSS or R is the better tool.
CHISQ.TEST(actual_range, expected_range) returns only the p-value, and it does not build the expected counts for you. Compute each expected count as the row total times the column total divided by the grand total, put them in a matching table, then point CHISQ.TEST at the observed and expected ranges. To report the chi-square statistic itself, sum (observed minus expected) squared divided by expected, or use CHISQ.INV.RT with the p-value and degrees of freedom.
Yes. Send your data file and the assignment and a statistician runs the analysis in Excel, reads the correct output cells and interprets it in full, before your deadline, with free revisions. Where Excel is the wrong tool for the design, we run it in SPSS or R instead.
Every task gets a custom quote based on the analysis, length and your deadline. Send the details and you see a free, no-obligation price before you pay anything.
Hand it to a statistician and get accurate output and a clear interpretation before your deadline.
Get my free quote →