CROSSTABULATIONS AND PROPORTIONS.
January 26, 2026
The Crosstabs describe the relationship between two categorical variables.
A frequency table describes one categorical variable. The “Crosstabs” (Cross-tabulation) makes a numerical summary of two variables and displays the information in a special type of table.
The cell contents of each combination of categories include the frequency, a percentage in proportion to the row total, and another percentage in proportion to the column total.
A hypothetical dataset contains sales revenue by continent and product type.
| Case | Continent | Product | Revenue |
|---|---|---|---|
| 1 | Australia | Product B | 930.00 |
| 2 | Africa | Product A | 584.00 |
| 3 | South America | Product A | 508.00 |
| 4 | North America | Product A | 1048.00 |
| 5 | Australia | Product A | 963.00 |
| 6 | North America | Product A | 1364.00 |
| 7 | South America | Product A | 477.00 |
| 8 | North America | Product B | 858.00 |
| 9 | Asia | Product C | 795.00 |
| 10 | Asia | Product C | 936.00 |
| 11 | South America | Product B | 507.00 |
| 12 | Europe | Product B | 855.00 |
| 13 | Europe | Product A | 930.00 |
| 14 | Australia | Product A | 1036.00 |
| 15 | Europe | Product A | 609.00 |
| 16 | Asia | Product C | 780.00 |
| 17 | Australia | Product A | 623.00 |
| 18 | Europe | Product A | 691.00 |
| 19 | Australia | Product A | 1081.00 |
| 20 | Europe | Product B | 497.00 |
One way to do this exercise is to 1 input everything just as it is, or we 2 first create value-labels and then input numbers (codes) into the nominal variables. You can generate crosstabs whichever way you choose.
Method 1. Create the variables (Continent, Product, and Revenue) directly in the variable view, and input the sample data into the data view.
| Name | Type | Values | Missing | Measure | |
|---|---|---|---|---|---|
| 1 | Continent | String | None | None | Nominal |
| 2 | Product | String | None | None | Nominal |
| 3 | Revenue | Numeric | None | None | Scale |
Method 2. Create the variables (Continent, Product, and Revenue) directly in the variable view, create value-labels, and then input numbers (codes) into the nominal variables.
| Name | Type | Values | Missing | Measure | |
|---|---|---|---|---|---|
| 1 | Continent | Numeric | {1, Africa}... | None | Nominal |
| 2 | Product | Numeric | {1, Product A} | None | Nominal |
| 3 | Revenue | Numeric | None | None | Scale |
Value-labels: Continent
- 1 = Africa
- 2 = Asia
- 3 = Australia
- 4 = Europe
- 5 = North America
- 6 = South America
Value-labels: Product
- 1 = Product A
- 2 = Product B
- 3 = Product C
Run SPSS Crosstabs…
- Analyze → Descriptive Statistics → Crosstabs…
- Move "Continent" to Row(s), and "Product" to Column(s)
- Cells: Counts → Select Observed
- Continue, OK
Once the data are cross-tabulated with SPSS, the output looks like the following table:
- Out of all 5 cases (row total), the proportion of product A in Australia = 4/5 = 0.8
- Out of all 5 cases (row total), the proportion of product B in Australia = 1/5 = 0.2
- The proportion of product A in Australia, out of all product A (column total) = 4/12=0.3
- Out of all 20 cases (column total), the proportion of product A = 12/20=0.6
- Out of all 20 cases, the proportion of product A in Australia = 5/20 =0.25
Row Percentages
- Analyze → Descriptive Statistics → Crosstabs…
- Move "Continent" to Row(s), and "Product" to Column(s)
- Cells: Percentages → Select (Row)
- Continue, OK
Home | Contact: info@gazersdigest.com
©2020-2026 Gazer’s Digest is an information sharing platform. Gazer’s Digest includes topics, thought to be useful for the readers.