How to: Unpaired t test from averaged data

The unpaired t test compares the means of two unmatched groups, assuming that the values follow a Gaussian distribution. This page gives detailed instructions for entering averaged data. Look elsewhere if you want to enter raw data.

1. Enter data

From the Welcome (or New Table and graph) dialog, choose the Grouped tab. (The t test is usually done from data tables formatted for column data, but Prism doesn't let you create column tables with subcolumns. Instead, create a Grouped table and enter data on one row).

Choose an interleaved bar graph, and choose to enter the data as Mean, SD and N (or as Mean, SEM and N).

Enter the data all on one row. Because there is only one row, the data really only has one grouping variable even though entered on a table formatted for grouped data.

Stop. If you want to compare three or more groups, don't use t tests repeatedly. Instead, use one-way ANOVA followed by multiple comparison post tests.

2. Choose the unpaired t test

1.	From the data table, click on the toolbar.

2.	Choose t tests from the list of Column analyses.

On the t test dialog, choose the unpaired t test. Choose the Welch's correction if you don't want to assume the two sets of data are sampled from populations with equal variances, and you are willing to accept the loss of power that comes with that choice. That choice is used rarely, so don't check it unless you are quite sure.

4.	Choose a one- or two-tail P value. If in doubt, choose a two-tail P value.

3. Review the results

The t test investigates the likelihood that the difference between the means of the two groups could have been caused by chance. So the most important results are the 95% confidence interval for that difference and the P value.

Learn more about interpreting the results of a t test.

Before accepting the results, review the analysis checklist.

4. Polish the graph

•	Be sure to mention on the figure, figure legend or methods section whether the error bars represent SD or SEM (what's the difference?).

•

To add the asterisks representing significance level copy from the results table and paste onto the graph. This creates a live link, so if you edit or replace the data, the number of asterisks may change (or change to 'ns'). Use the drawing tool to add the line below the asterisks, then right-click and set the arrow heads to "half tick down'.

•	To make your graph simple to understand, we strongly recommend avoiding log axes, starting the Y axis at any value other than zero, or having a discontinuous Y axis.