Create Tables with Multiple Variables

Row variables are those dictionary items or value sets which appear in the rows of the table. Column variables are those dictionary items or value sets which appear in the columns of the table. Every table must have at least one row or column variable specified; any given subtable can have a maximum of two row variables and two column variables. The number and disposition of row and column variables will affect the type of table generated.

• 1 Row/0 Column Variables

If a table consists of one row variable and no column variables, the system will produce a table with the variable's categories as rows and frequency counts as the only column.

• 0 Row/1 Column Variable

If a table consists of no row variables and one column variable, the system will produce a table with the tabulation categories in the columns and totals in the single row of the table.

• 1 Row/1 Column Variable

If a table or subtable consists of one row variable and one column variable, the system will produce a normal cross-tabulation, with the tabulation categories of each variable in row or column, as appropriate. By default, totals will appear in the left-most column and in the top-most row (it is possible, however, to change the position of the total).

• 2 Row/Column Variables

When a table or subtable is designed with two variables or value sets in the row and/or column, one of each pair is considered to be the independent [major] variable, and the other is considered to be the dependent [minor] variable. The tabulation categories of the dependent variable appear nested within the categories of the independent variable. Totals for a dependent variable appear as the topmost row or left-most column within each tabulation category of its independent variable.

Because there are effectively no limits on the number of rows and columns in a table, the combination of two variables/value sets can produce tabulations which will be extremely difficult to view and to understand. Users should give careful thought to the placement of variables and value sets in rows and columns, particularly when one or more of the items has a large number of tabulation categories. It is almost always easier to manipulate tables with large numbers of rows than those with the same number of columns.

Whenever the area processing function is invoked for a table set, the area levels are included as additional row categories within which the other row variables are displayed.

Subtables are created by adding variables to the bottom of the existing rows or to the right of the existing columns. When a subtable is created it is usually outlined in color to designate this fact. This is an example.

Notice these are like two separate tables side-by-side; 1) Marital Status by Sex and 2) Marital Status by Literacy. There is a total of 6 columns (3 for Sex plus 3 for Literacy). Note also the wording in the generated title.

Compare the above to the following:

This is a single table (1 row variable and two column variables) with a total of 9 columns (3 for Sex times 3 for Literacy). Note the wording in the generated title.