Cross Table

The Burt table is the basis of the multiple correspondence analysis. It is a partitioned symmetric table of all pairs of categorical variables. The diagonal partitions are diagonal matrices (a cross-table of a variable with itself). The off-diagonal partitions are ordinary contingency tables. When you select multiple Y, Response columns with no X, Factor columns, the Burt table is created. If you select any X, Factor columns, a traditional contingency table is created instead of a Burt table.

The red triangle menu for the Burt or contingency table contains options of statistics to display in the table.

Count

Cell frequency, margin total frequencies, and grand total (total sample size). This appears by default.

Total %

Percent of cell counts and margin totals to the grand total. This appears by default.

Cell Chi Square

Chi-square values computed for each cell as (O - E)2 / E.

Col %

Percent of each cell count to its column total.

Row %

Percent of each cell count to its row total.

Expected

Expected frequency (E) of each cell under the assumption of independence. Computed as the product of the corresponding row total and column total divided by the grand total.

Deviation

Observed cell frequency (O) minus the expected cell frequency (E).

Col Cum

Cumulative column total.

Col Cum %

Cumulative column percentage.

Row Cum

Cumulative row total.

Row Cum %

Cumulative row percentage.

Make Into Data Table

Creates one data table for each statistic shown in the table.