Polsinelli at Work |  Labor & Employment Blog

Employers, whether large or small, face an ever-growing web of workplace regulations and potential entanglements with employees. With employment litigation and advocacy experience as our strength, preventing legal problems from arising is our goal. Our Labor & Employment attorneys advise management on complex employee relations and workplace issues. 20 offices; 800+ attorneys. 

PolsinelliAtWork.com was recently recognized as one of the top employment blogs in the nation by Feedspot.
 


Three Considerations for Using Big Data in Hiring Decisions

By Gillian McKean Bidgood

With job candidates posting extensive information on social media and other information available on the Internet, technologists are developing ways to mine and use that data in the hiring process. This field (sometimes referred to as “people analytics”) is marketed as full of promise, including the possibility of identifying unrealized potential, increasing diversity, reducing turnover, improving employee satisfaction, and improving the company and individual performance. However, for employers inclined to embrace people analytics, there are a number of employment law-related issues to consider.

1.    Statistics Are Not Inherently Objective

People analytics may help reduce the subjective assessments that are inherent in the interview process. However, to create a tool to predict success on the job or identify “desirable” traits for job applicants, an employer must first define what makes an employee successful or the traits that are desirable. Typically, the logical starting point is an employer’s current workforce. The current workforce may not include “successful” employees as the employer would now define the job or the “desirable” traits for the job going forward. As such, tools based on the current workforce may perpetuate the issues found in the current work environment. Employers may wish to consider whether the data used or the tool itself should be adjusted to counter those tendencies. In addition, employers should consider whether augmenting their own data with data from outside of the company could improve the objectivity of the data.

2.    Correlation Is Not the Same as Causation

Analysis of data about existing employees or workers in the industry will likely reveal many interesting connections. It is easy at first to erroneously assume that a connection is causal. For example, even if there is a correlation between playing team sports in school and ultimately succeeding on the job, participating in team sports may not be the reason that the employees are ultimately successful. If the technologist or employer focuses more on the measurable indicator (team sports) than what the measurable indicator reflects (e.g., time management), the predictive value of the tool may suffer and the tool may have unintended effects. Accordingly, technologists and employers should not limit their thinking to finding correlations and should consider what the correlations mean about the applicant or employee’s skills and abilities. 

3.    Technologists, HR and Legal Teams Should Partner on People Analytics

An employer might save time and reduce legal risk by having technologists develop or implement a people analytics tool alongside the employer’s HR and legal teams. The HR and legal teams can help the technologists avoid creating or implementing a tool that results in discrimination or violates other laws, such as privacy laws and the Fair Credit Reporting Act.