Less bias bigger gap?

From Online tests are the latest gateway to landing a new job by By Sarah Halzack. This holds great promise in terms of productivity, fairness, and human capital development. The promise is a cheaper recruiting process, more accurate selection, better employee performance and reduced employee churn. But the transition will be hard as many politically popular bromides are either circumvented or forced to confront reality.

If you’re applying for a job as a customer service representative at T-Mobile, you’re bound to encounter Jason Easton, a cranky mock customer who has been on hold for nearly an hour.

“Ah! It’s about time,” he says, demanding to know why his bill has gone up.

As Easton rattles off his name and phone number, you’ll have to quickly pull up his account, help him with his bill and determine whether he’s eligible for a $30 credit that he wants.

T-Mobile asks job applicants to take this test before inviting them for an interview because the company has found powerful correlations between the online assessments and success on the job. High scorers tend to resolve customer calls about 25 seconds faster than those who receive low scores. That means they can handle one more call a day and about 250 more a year.

At T-Mobile and legions of other companies, Web-based tests have become a key gateway to landing a job, a potent screening tool that can effectively bump a résumé to the top or bottom of a manager’s pile.

Some of the benefits of online testing also represent risk.

Test makers say their offerings bring a consistency and objectivity to a process that can sharply improve the odds of hiring the right person. But in a highly competitive job market in a tepid economic recovery, the increased use of online testing could mean that workers who aren’t digitally savvy or lack Web access might face one more hurdle in getting a job.

“Assessments are right more often than they’re wrong,” said Elliot Clark, chief executive of SharedXpertise Media, a firm that puts on conferences for the human resources industry. “But like anything else, when you do a prediction, the forecast has a percentage of accuracy. The issue is: What percentage of people get screened out that should have gotten a shot?”

The authors hold out what I suspect might be a misguided hope.

Providers say the tests hold the promise of leveling the playing field for job applicants by removing the chance of bias that comes with a traditional résumé screening. The tests can’t distinguish, for example, if a candidate didn’t attend a top-tier college, is currently unemployed or is a woman or minority.

Bias is a hard to quantify component in human interactions and by disintermediating it, on-line testing is an advance. But by relying strictly on objective empirical quantifications, there is likely to be disparate impact of some sort, in some instances possibly extreme. If human bias is a smaller factor than the real world empirical objective skill gap, then it is possible that on-line tests might lead to greater disparity rather than less.