From the course: Ethics and Law in Data Analytics

Employment and policy

From the course: Ethics and Law in Data Analytics

Start my 1-month free trial

Employment and policy

“

- New hiring practices use algorithms in place of industrial psychologists in hiring recruiters. It makes sense that an algorithm could be designed to take bias out of hiring. Such as the like me bias that leads people to hire people who are like themselves over others not like themselves. But there is here a risk that since the algorithm is designed by humans it will somehow incorporate human bias. An example given in the 2016 White House Big Data Report is this. What if a data point used in the algorithm is the age that the job candidate become interested in computing? There are cultural messages and assumptions that associate computing with boys more often than girls. And use of this data point could promote more male hires than female hires. It could skew hiring in favor or men over women. Here we see potential discard impact where the use of what seems like a neutral data point, the age at which the person became interested in computers can have a discriminatory impact on a protected class, women. Existing employment discrimination laws such as the Civil Rights Act might protect victims of this type of discrimination. But again because of the transparency issue, unknown data points and unknown processes, it is nearly impossible for a job applicant to realize she has been discriminated against.

Contents