Written and maintained by CASRAI Editorial Board
Last updated
In Minitab, one-way ANOVA is run from Stat > ANOVA > One-Way, and two-way ANOVA is run through Stat > ANOVA > General Linear Model > Fit General Linear Model with the interaction added in the Model dialog. Minitab’s own documentation sends designs with two or more fixed categorical factors to the general linear model, not to a separate two-way menu item. This guide covers both routes, then shows how Minitab’s output is read.
What it is used for in research
ANOVA tests whether the population means of two or more groups differ. Minitab’s help states the condition for the one-way version plainly: use it when you have a categorical factor and a continuous response. If the test says at least one group differs, a comparisons analysis then shows which pairs differ. For the statistics behind the test, see CASRAI’s guide to analysis of variance and the dictionary entry ANOVA (analysis of variance).
Which Minitab analysis fits which design
Minitab’s one-way ANOVA overview lists the alternatives to use when the design does not fit:
| Your design | Minitab analysis named in its help |
|---|---|
| One categorical factor, continuous response | One-Way ANOVA |
| Two or more fixed categorical factors | Fit General Linear Model |
| Random factors | Fit Mixed Effects Model |
| One continuous predictor | Fitted Line Plot |
| More than one predictor | Fit Regression Model |
If you are still choosing a test, start with how to choose a statistical test.
Install or access (official source only)
Minitab is obtained from the maker’s own product page, minitab.com/en-us/products/minitab. Choose the delivery form there. The menu paths below are the ones given in Minitab’s current support documentation.
Step-by-step task
One-way ANOVA in Minitab
- Choose Stat > ANOVA > One-Way.
- Enter the continuous response and the categorical factor in the dialog.
- To control the output, click Results. Simple tables displays the simpler version of all tables; Expanded tables adds the expanded ANOVA, Model summary and Multiple comparisons tables.
- Click OK in each dialog.
Two-way ANOVA in Minitab
Minitab’s worked procedure uses a response A and two factors B and C.
- Choose Stat > ANOVA > General Linear Model > Fit General Linear Model.
- In Responses, enter A. In Factors, enter B and C.
- Click Model. Select both factors in Factors and covariates.
- To the right of Interactions through order, choose 2 and click Add. This adds the B*C interaction term.
- Click OK in each dialog.
With the interaction in the model, a significant interaction means the effect of one factor depends on the level of the other, so read it before the main effects. CASRAI’s guide to simple effects after a significant interaction covers that step. The same two-factor design in another package is in two-way ANOVA in SPSS.
Output and how to read it
One-way output is organised as Method, Factor information, Analysis of variance, Model summary and Means. Read it in this order, using the example in Minitab’s key-results page.
- p-value against the significance level. In Minitab’s example the p-value is 0.004, below the significance level of 0.05, so the null hypothesis is rejected. If p is at or below alpha, some means differ significantly; if it is above, there is not enough evidence of a difference. See p-value.
- R-sq and R-sq(pred). R-sq is the percentage of variation in the response explained by the model (47.44% in the example). A high R-sq alone does not show that the assumptions are met, and an R-sq(pred) well below R-sq (24.32% in the example) can signal overfitting.
- Grouping information. Groups that do not share a letter are significantly different.
- Tukey confidence intervals. Intervals that do not contain zero indicate a statistically significant mean difference. Multiple comparisons inflate the overall error rate, so the simultaneous confidence level matters. For the method itself see Tukey HSD test.
Troubleshooting
- The p-value is above your significance level. The data give not enough evidence that the means differ; do not run pairwise comparisons expecting a different answer.
- You need to know which groups differ. A significant overall test only says that at least one group differs. Turn on Expanded tables in Results to display the Multiple comparisons table.
- Your factor or response is the wrong type. One-Way ANOVA needs a categorical factor and a continuous response. For other designs use the alternatives in the table above.
- You expected a separate Two-Way menu item. Use Fit General Linear Model with the interaction added, as in the steps above.
Free alternatives
| Tool | CASRAI guide |
|---|---|
| R | Running ANOVA in R |
| Python (statsmodels) | ANOVA in Python with statsmodels |
| JASP | JASP statistical software |
For a wider list, see free statistical software for research. For more on what Minitab does beyond ANOVA, see what Minitab is used for.
FAQ
How do I run a one-way ANOVA in Minitab?
Choose Stat > ANOVA > One-Way, enter the response and the factor, and click OK. It requires a categorical factor and a continuous response.
How do I run a two-way ANOVA in Minitab?
Choose Stat > ANOVA > General Linear Model > Fit General Linear Model, enter the response and both factors, open Model, and add the interaction with Interactions through order set to 2.
How do I tell which groups differ after a significant ANOVA?
Use the comparisons output. Groups that do not share a letter in the grouping information are significantly different, and Tukey intervals that exclude zero mark significant differences.
What if I have random factors?
Minitab’s help points to Fit Mixed Effects Model for random factors.
Sources (official docs)
- Minitab Support: One-Way ANOVA overview, checked 2026-10-29.
- Minitab Support: select the results to display, checked 2026-10-29.
- Minitab Support: One-Way ANOVA key results, checked 2026-10-29.
- Minitab Support: perform a two-way ANOVA, checked 2026-10-29.
- Minitab product page, checked 2026-10-29.
