Bias measurement
Understand how to properly measure unwanted bias in offering products to
your customers to comply with your company policies and regulations.
The bias
threshold measurement depends on the type of field that you selected. Depending on your
settings, you can select the rate ratio or Gini coefficient.
- Rate ratio
- Use this ratio to determine bias for categorical fields by comparing the
number of customers who were selected for an action to those not
selected for an action, and correlating that to the selected bias field.
For example, the rate ratio that is represented in the following table
indicates that actions are sent more often to male rather than female
customers:
A rate ratio of 1 represents perfect distribution equality. You can select a warning threshold between 0 (warn if any bias is detected) and 0.7 (warn only if very high bias is detected). You can also choose to ignore this bias field for a particular issue in your business structure.
Example rate ratio Female customers Male customers selected for the action 500 1000 not selected for the action 20000 18000 rate ratio [500 / (500+20,000) ] / [1000 /(1000+18000] = 0.46 [1000 / (1000+18,000) ] / [500 /(20,000+500] = 2.16 - Gini coefficient
- Use the Gini coefficient calculate bias for numerical fields. This is a method of measuring the statistical inequality of value distribution, for example, the distribution of actions to customers based on their age. A Gini coefficient of 0 represents perfect distribution equality. You can select a warning threshold between 1 (warn if any bias is detected) and 0.50 - 2.00 (warn only if very high bias is detected). You can also choose to ignore this bias field for a particular issue in your business structure.