Skip to main content
v2026.11,986 entries · CC-BY 4.0

Prism Statistics: What GraphPad Prism Does and How to Run Your First Analysis

What GraphPad Prism does for research statistics, how to run a first analysis step by step, read the output and free alternatives.

Written and maintained by CASRAI Editorial Board

Published

On this page

Prism statistics refers to the analyses built into GraphPad Prism, a statistics and graphing application for scientific research that runs on Windows and Mac. Prism lets you enter data in a spreadsheet-style table, choose an analysis from the Analyze menu or Analyze button, and read the results and graph without writing code. This guide explains what it is used for, how to install it from the official source, how to run a first analysis step by step, how to read the output, what to do when something goes wrong, and which free programs do comparable work.

What GraphPad Prism statistics is used for in research

GraphPad describes Prism on its features page as ‘the preferred analysis and graphing solution purpose-built for scientific research’ and says it is available on Mac and Windows. In practice it is used by bench and clinical researchers who want statistical tests, curve fitting and publication graphs in a single application. The word ‘Prism’ alone is ambiguous, so on this page it always means GraphPad Prism.

Prism statistics software: the analyses on offer

The features page lists the following families of analysis (read 2026-11-14):

  • Group comparisons: paired and unpaired t tests; one-, two- and three-way ANOVA; repeated-measures ANOVA; mixed-effects models; nested tests; and the nonparametric Mann-Whitney, Wilcoxon, Kruskal-Wallis and Friedman tests.
  • Categorical data: Fisher’s exact test, chi-square and Kolmogorov-Smirnov tests.
  • Multiple comparison corrections: Tukey, Dunnett, Bonferroni, Holm-Sidak and false discovery rate (FDR) approaches.
  • Nonlinear regression: more than 100 built-in equations or a custom equation, global fitting, robust regression, outlier detection, and model comparison by the extra sum-of-squares F test or AICc, with confidence and prediction bands.
  • Survival analysis: Kaplan-Meier curves with the log-rank test, and Cox proportional hazards regression.
  • Other regression: simple and multiple linear, logistic, Poisson (generalised linear model) and Deming regression.
  • Multivariate and clinical-lab tools: principal component analysis, Bland-Altman analysis and ROC curves.
  • Descriptive work: descriptive statistics, normality tests, outlier tests (Grubbs and ROUT), power and sample size, simulations and data transformation.

Prism statistical software versus a programming language

The features page states ‘No coding required!’ in the regression section, where Prism automatically encodes categorical variables for multiple linear and multiple logistic regression. That is the main design difference from a scripting environment such as R: analyses are chosen from menus and dialogs, and each result stays linked to its data table. For a related question that many researchers ask next, see how to read significance on a figure in showing a P value on a Prism graph.

Install or access GraphPad Prism from the official source

Download Prism only from graphpad.com, the vendor’s own site. The system requirements published in the Prism 11 user guide (read 2026-11-14) are:

  • Windows: 64-bit Windows 10 or 11; 32-bit editions are not supported. An x64-compatible CPU is required; there is no native ARM build, but it can run on ARM devices through x64 emulation on Windows 11. About 170 MB of disk space. Microsoft Edge WebView2 is needed to display the Welcome dialog, and the installer adds it if it is missing.
  • Windows memory: 2 GB of RAM for files up to 2 million data cells, 4 GB for 2 to 8 million cells, 8 GB for 8 to 16 million cells and 16 GB above 16 million cells.
  • macOS: macOS 12 (Monterey) or later, as a Universal Binary that runs natively on Apple silicon and Intel Macs. Support for macOS 12 ends with version 11.3.0, which requires macOS 13 (Ventura) or later. About 530 MB of disk space.
  • Licence check: activation needs an internet connection. Prism checks the licence at every launch and every 24 hours if left open, and the check must succeed once every 30 days or within 20 attempts, whichever comes first.

Licence types are a separate question; the existing guide on the GraphPad Prism student licence covers that route.

Step-by-step: run your first Prism statistics analysis

The steps below follow the guided tour in the current Prism user guide (read 2026-11-14), which analyses sample data with an unpaired t test. The same pattern applies to most analyses: pick the right data table, enter data, analyse, check, graph.

  1. Start Prism. The Welcome (File..New) dialog appears, with one tab for each of the eight kinds of data table.
  2. Click the Column tab, choose to use sample data, and select the sample data for an unpaired t test. This fills the table with data only; you choose the analysis in the next step.
  3. Read the floating yellow note on the data table. It explains the data format and links to online help.
  4. With the data table showing, click Analyze and choose t test. A t test shortcut button above the Analyze button does the same thing.
  5. On the first tab of the dialog, confirm that the design is unpaired, that you assume a Gaussian distribution, and that you want an unpaired t test. Accept the defaults on the second and third tabs. Click Learn beside any option for an explanation.
  6. Review the results sheet. Click the Analysis checklist button to see questions that help you interpret the output.
  7. To change the test, click the Results button in the upper left corner. This opens the Analysis Parameters dialog, where you can switch to a paired or nonparametric test.
  8. In the navigator on the left, under graphs, click the graph named for the unpaired t test data. The Change Graph Type dialog opens.
  9. In the Draw section of the toolbar, click Pairwise Comparisons to add the t test result to the graph. Click it again to format how it looks.

For the choices and reporting around this particular test, see running a t test in GraphPad Prism.

Two ways to choose any other analysis

The user guide on choosing an analysis describes two routes. The Analyze menu, which sits between the Insert and Change menus, groups analyses by objective, and choosing one opens the Analyze Data dialog with that analysis already selected. The Analyze button on the toolbar opens the Analyze Data dialog directly, where analyses are grouped by the type of data table they are usually used with. In that dialog the top dropdown should usually stay on ‘Built-in analysis’ (the alternative is a saved method), the search bar filters the list by name, and the right half limits the analysis to selected data sets. Prism ignores any selection you made before opening the dialog, so you do not need to select data first.

Chaining analyses and project structure

A Prism project is one file that can hold up to 500 tables, graphs, analyses and layouts. From any results table with green grid lines you can click the Analyze button to analyse the results themselves, for example transforming data and then running nonlinear regression on the transformed values (Prism user guide, ‘How to begin’, read 2026-11-14).

Output and how to read it

Each analysis creates a results sheet. The first read should be of the choices you made: the Analysis Parameters dialog, reachable from the Results button, records whether the test was paired or unpaired and parametric or nonparametric. Then use the Analysis checklist button in the Interpret section of the toolbar to work through the interpretation questions Prism provides for that analysis. The graph is linked to the data and results, so changing the data table updates both.

Two outputs in the tour are worth noting for reporting. First, Pairwise Comparisons places the test result on the graph itself. Second, the estimation plot, found in the navigator under graphs, shows the effect estimate and its 95% confidence interval instead of a P value. The data table below lists which kinds of data table suit which analyses. When you run many comparisons, read the separate guidance on Bonferroni correction and multiple comparisons before choosing a correction, and note that Prism lists Tukey, Dunnett, Bonferroni, Holm-Sidak and FDR among its options.

Prism also ships a companion pair of manuals, the Prism Statistics Guide and the Curve Fitting Guide, which explain the statistical reasoning behind the analyses. The statistics guide opens its analysis chapter by saying to start from a data table, graph or green results table and click the Analyze button. If your next step is a quick P value outside Prism, the guide on the GraphPad P value calculator covers that route.

Troubleshooting

  • The installer or app will not run on an older computer. Prism 11 for Windows needs 64-bit Windows 10 or 11, and macOS builds from version 11.3.0 need macOS 13 or later. 32-bit Windows is not supported.
  • The Welcome dialog does not show. It depends on Microsoft Edge WebView2 on Windows; the installer adds it when missing, so reinstalling with the official installer is the documented route.
  • A licence check fails. Prism needs an internet connection for activation and checks at every launch and every 24 hours while open. A check must succeed at least once every 30 days or within 20 attempts, so a computer kept offline for a long time will eventually need to reconnect.
  • You cannot find an analysis. Open the Analyze Data dialog and type part of its name, such as ‘Nonlinear’, in the search bar. Remember that the list is grouped by data table type, so the analysis may sit under a different table than the one you expected.
  • You picked the wrong test. Click the Results button to open Analysis Parameters and switch between paired, unpaired and nonparametric versions without re-entering data.
  • An option is unclear. Click Learn in the dialog to open its help topic.

Free alternatives to Prism statistics

These programs cover overlapping ground. Descriptions are taken from each project’s own site (read 2026-11-14); none replicates Prism’s curve-fitting library or graph layout system in an identical way, so check that the analyses you need are available before switching.

Program How its own site describes it Cost and licence stated Interface
R ‘A language and environment for statistical computing and graphics’; compiles and runs on UNIX platforms, FreeBSD, Linux, Windows and MacOS Free software under the GNU General Public License Code-based
jamovi ‘A free statistical platform, built by the community’; spreadsheet-style, runs on R, desktop and cloud Free; the site accepts donations Spreadsheet and menus
JASP A free, open-source, cross-platform program with a graphical interface, supported by the University of Amsterdam; frequentist and Bayesian versions of standard procedures Free and open source Graphical menus

Frequently asked questions

What is Prism statistics?

It is the set of statistical analyses and graphing tools in GraphPad Prism, an application for scientific research on Windows and Mac. It covers t tests, ANOVA, nonparametric tests, nonlinear regression, survival analysis and more, chosen from menus rather than code.

Is Prism statistical software only for biology?

GraphPad describes Prism as built for scientific research and lists clinical-lab tools such as ROC curves and Bland-Altman analysis alongside general tests. The analysis list is not limited to one discipline.

Does Prism need coding?

No. GraphPad’s features page says ‘No coding required!’ in its regression section, and the user guide’s tour runs a complete analysis from menus and buttons.

Which data table should I choose?

Choose it by the design of the experiment, because the table determines which analyses and graphs are available. The data table below pairs six table kinds with their typical analyses.

How do I change a test after running it?

Click the Results button at the upper left of the results sheet to open Analysis Parameters, then switch to a paired or nonparametric version.

Sources

  • GraphPad Software. Prism features. https://www.graphpad.com/features (checked 2026-11-14)
  • GraphPad Software. Prism 11 user guide, system requirements. https://www.graphpad.com/guides/prism/latest/user-guide/system_requiremens.htm (checked 2026-11-14)
  • GraphPad Software. Prism user guide, A brief tour of Prism. https://graphpad.com/guides/prism/latest/user-guide/a_brief_tour_of_prism.htm (checked 2026-11-14)
  • GraphPad Software. Prism user guide, Choosing an analysis. https://www.graphpad.com/guides/prism/latest/user-guide/using_choosing_an_analysis.htm (checked 2026-11-14)
  • GraphPad Software. Prism user guide, How to begin. https://graphpad.com/guides/prism/latest/user-guide/how_to_begin.htm (checked 2026-11-14)
  • GraphPad Software. Prism user guide, Statistical analyses with Prism. https://graphpad.com/guides/prism/latest/statistics/stat_statistical_analyses_with_pris.htm (checked 2026-11-14)
  • GraphPad Software. The five kinds of data tables in Prism (knowledgebase article 1701). https://www.graphpad.com/support/faq/the-five-kinds-of-data-tables-in-prism/ (checked 2026-11-14)
  • The R Project. About R. https://www.r-project.org/about.html (checked 2026-11-14)
  • jamovi. About. https://www.jamovi.org/about.html (checked 2026-11-14)
  • JASP. https://jasp-stats.org/ (checked 2026-11-14)

Data: GraphPad Prism data table kinds and the analyses they typically support

Compiled from GraphPad's Prism user guide and knowledgebase article on data table kinds: for six of the eight Prism table kinds, what each table holds and which analyses it is typically used for. Prism has eight kinds in total; only the six confirmed in the sources read are listed.

GraphPad Prism data table kinds and the analyses they typically support
Data table kindWhat it holdsTypical analyses
XYEvery point defined by an X value and a Y valueLinear and nonlinear regression
ColumnGroups defined by one scheme (for example control versus treated), one group per columnt tests, one-way ANOVA
GroupedLike Column tables but for two grouping variablesTwo-way ANOVA
ContingencyCounts of subjects or observations falling into categories set by the rows and columnsChi-square, Fisher's exact test
SurvivalTime to event with outcome coded as death (1) or censored (0)Kaplan-Meier survival analysis, log-rank test
Multiple variablesEach row is a case and each column is a variable, the layout most statistics programs useCox proportional hazards regression

Sources

Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about Prism Statistics: What GraphPad Prism Does and How to Run Your First Analysis

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →