A multiple variable table is arranged the same way most statistics programs organize data. Each row is a "case" (experiment, animal, etc.). Each column is a different variable. There are no subcolumns in multiple variable tables.

Analyses performed on multiple variable data

•Correlation matrix

•Multiple linear regression

•Extract and rearrange portions of the data onto a new table

•Transform and select

•Identify outliers

•Descriptive statistics

Example of a multiple variable table

The example below shows the first few rows of data from a clinical trial about diabetes. Each row represents one person studied. Each column is a different variable measured.

Prism does not automatically graph these kinds of tables, because there is too much data.

Prism requires that you enter numbers only. If one column encodes sex, you must enter codes (say 0 for female and 1 for male). You can't enter "male" or "female" directly on a Prism table.

.