How does GraphPad Prism compare with Microsoft Excel?

Microsoft Excel is a general business application that is very good at storing and manipulating large amounts of data. Excel does not make quality scientific graphs, and can perform only a few statistical analyses. While it is possible to coax Excel to fit curves, it isn't easy and the results are not complete (no standard errors or confidence intervals of the parameters). Likewise, Excel does not have the flexibility to create the presentation-quality graphs most scientists require.

Prism is designed specifically for research scientists and makes it easy to perform the analyses scientists need, including t-tests, one- and two-way ANOVA, parametric and non-parametric tests, curve fitting, survival analyses and more. Prism also allows users to create and customize presentation-quality graphs, with features such as automatic error-bars, log axes, offset and discontinuous axes, and custom ticks and grid-lines. Prism can create many graphs that Excel simply cannot, including scatter graphs, box-and-whiskers plots, and survival curves. Prism also provides tools to automate repeated analyses and graphing. Prism's graphs and layouts can be exported directly to PowerPoint, or as high-resolution Tiff, PDF, or EPS files.

Many, perhaps most, Prism users also use Excel and copy and paste data into Prism for analysis and graphing. With the Mac version of Prism, simply paste the values into Prism. With the Windows version, you can choose to paste a link to the original Excel file, or you can paste an embedded Excel worksheet within Prism. In either case, you won't be able to edit those values in Prism (without breaking the link). Instead, you can double-click on the data block in Prism to return to Excel. Edit the data within Excel (or edit other values which contribute to the calculation of the values you pasted) then close the Excel worksheet. The data within Prism are automatically updated.

Here is a partial list of features missing from Excel, but present in Prism:

• No automatic error bars on XY graphs--user can write functions to compute error bars and to graph them on XY graphs, but with difficulty
• No error bars on bar graphs
• No nonparametric statistics
• No post tests following one-way ANOVA
• Two-way ANOVA does not allow missing values
• No repeated-measures ANOVA (one- or two-way)
• No normality testing
• No survival analysis
• No ROC curves
• No Deming regression
• No Fishers exact test
• Nonlinear regression is very difficult, with incomplete results.
• No standard errors or confidence intervals from nonlinear regression
• No interpolation from a standard curve following linear or nonlinear regression
• No statistical teaching on Help screens
• No column scatter graphs
• No analysis of survival curves
• No discontinuous axes
• No offset axes
• No custom (additional) ticks and gridlines
• No box-and-whiskers plots
• User can't include data and results as table inset within a graph
• No easy way to arrange several graphs for printing or export
• User can't export graphs in other formats (tiff, jpg, PDF, EPS, etc.)