4.3 CONTINGENCY TABLES
To help quantify the relationship between a categorical predictor and the target, we can construct a contingency table, which is a cross‐tabulation of the two variables, and contains a cell for every combination of variable values (that is, for every contingency). Figure 4.4 contains a contingency table of previous_outcome with response. Note that the usual practice is to have the target variable representing the rows, with the predictor representing the columns. For EDA, it is also helpful to include the column percentages. Figure 4.5 contains the table with column percentages. Most customers had no previous marketing campaign (nonexistent), so note that 21,176 of these responded no while 2034 responded yes. Overall, note that the proportion of yes response is only 13.9% for failure and only 8.8% for nonexistent, but a very high 64% when the customer's previous marketing campaign was a success.
![No alt text required.](https://static.packt-cdn.com/products/9781119526810/graphics/images/c04f004.gif)
Figure 4.4 Contingency table from R of previous_outcome...