Significance Variables
Specify variables from the {link}{Emphasis}Input SAS Data Set that contain continuous significance measures or numeric binary indicators created from continuous significance measures.
Common continuous measures to use are t-statistics, -log10(p-values), or p-values. The significance variables specified here should all be of the same kind (for example. all t-statistics or all -log10(p-values)). Basic characteristics or uses of each are as follows:
| • | T-statistics are recommended for simultaneously testing both up- and down-regulation. |
| • | P-value-based measures only screen for absolute significance. |
| • | -Log10(p-values) are preferred to p-values for Cochran-Armitage and PAGE tests because they have more compatible sampling distributions. |
If binary indicators are specified here, you must specify a value in the {link}{Emphasis}Significance Variable Cutoff for Fisher Exact Tests field on the Analysis tab. For example, if you specify one or more binary indicators in which 0 represents non-significance and 1 represents significance, you must specify a value greater than 0 and less than 1 as your {link}{Emphasis}Significance Variable Cutoff for Fisher Exact Tests.
To Specify Significance Variables:
| 8 | Specify an {link}{Emphasis}Input SAS Data Set. |
The Available Variables field is populated with variables from the specified data set.
| 8 | Highlight a single variable, or hold down while left-clicking on multiple variables, from the Available Variables field. |
| 8 | Click  to add the selected variable(s) to the Significance Variables field. |
