Statistical testing assumptions

In general, Displayr conducts its statistical testing using nonparametric tests, which make relatively few assumptions (compared to other tests).  
You can modify how Displayr conducts its significance tests by selecting one or more objects and then in the Object Inspector go to Properties > APPEARANCE > Significance.  This allows you to choose between:

• Testing at the 0.05 level of significance with a false discovery rate multiple comparison correction applied to all the tests on the table.  
• Testing at the 0.05 level of significance (i.e., the 95% level of confidence).  This is the default.
• Testing at the 0.10 level of significance (i.e., the 90% level of confidence).

See also

• Comparing differences between columns.
• Interpreting the arrows and colors.