Data vs. Discretion in Hiring

When it comes to relying on either hiring managers' judgement or pre-hiring assessments in making hiring decisions, new research gives a slight edge to the assessments. But experts say nothing can replace a collaborative relationship between recruiters and hiring managers.

Thursday, December 17, 2015
Write To The Editor Reprints

Trust in your instinct has just been kicked in the gut again.

A recent study supports the data-driven argument that hiring managers shouldn't contradict pre-hiring assessment results. Published by National Bureau of Economic Research in Cambridge, Mass., the report highlights the evolving debate over how much human discretion there should be in the hiring process.

The results suggest that, if more discretion is associated with a lower duration of a new hire's job tenure, then potentially increasing discretion would not be beneficial to organizations, according to Mitchell Hoffman, of the University of Toronto Rotman School of Management, who co-authored the report with Lisa B. Kahn of Yale University School of Management and Danielle Li of Harvard Business School.

The study looked at some 691,352 applicants for low-skill service sector jobs at 15 firms; some 555 managers were involved in hiring nearly 300,000 employees. While job testing substantially improved the match quality of the hired workers -- they had 15-percent longer tenures (12 days over a 12-month period) than those hired without testing -- managers who overruled test recommendations more often hired workers with lower match quality, as measured by job tenure.

The authors also suggest that "managers exercise discretion because they are biased or have poor judgment, not because they are better informed. This implies that firms in our setting can further improve match quality by limiting managerial discretion and placing more weight on the test."

With more than three quarters of companies using some kind of employment test for external hires and another 12 percent planning on using something in the future, according to CEB, getting the best employees can be undermined if hiring managers are not given clear guidelines for when they can use their discretion to overrule an assessment result, industry watchers say.

The continuing triumph of data over discretion may be "a little uncomfortable emotionally for managers," says Ernie Paskey, partner and lead of assessment solutions for Aon Hewitt in Washington. Assessment data needs to be interpreted and presented to the manager in a meaningful way, he says, to help them make a quality hiring decision.

Ravin Jesuthasan, the Chicago-based global practice leader for Towers Watson's talent management practice and author of Transformational HR, says making biased errors is only human: "It's tough to overcome 20, 30 or 40 or 50 years of conditioned response."

However, it is increasingly critical for companies to not make a hiring mistake these days, says Jesuthasan. "In the past, we had much more of a luxury and tolerance for maybe getting it almost right," he says. "Now, a much more volatile economy and continually shrinking budget mean that you want to be 110-percent correct before embarking on a hire. I think that [need for security] is one of the reasons why it took so long for hiring to pick up after the economy started to recover."

However, the 12-day difference uncovered by the NBER research is not enough to sway Robin Erickson, vice president of Bersin by Deloitte's talent acquisition, engagement and retention research group in Chicago. While her company's research shows pre-hire assessment is helpful, she says the clear winner in 15 different factors "is a good, communicative relationship between recruiters and hiring managers -- and that requires humans in the process."

While the NBER study looked at the results of more traditional questionnaire-based assessments, there are plenty of new assessment tools -- particularly gamification, which can appeal to millennials because they grew up with the technology being used to assess them, says Barb Marder, a Baltimore-based senior partner in Mercer's talent business and team leader in the company's Innovation Hub.

Unlike multiple-choice tests, in which applicants can answer what they think the employer is looking for, games can't be "gamed" by applicants because their abstract nature hides the assessment goal, says Marder. The games can more accurately measure the candidate's decision-making processes, creativity, skills and more, all without cluing them in to bias their answers.

Newsletter Sign-Up:

HR Technology
Talent Management
HR Leadership
Inside HR Tech
Special Offers

Email Address

Privacy Policy

She says some companies have their high performers use an assessment tool and the employer's target benchmarks for new workers are then developed based on the high-pos' results.

To wit, Mercer recently invested in Pymetrics, a game-based career-search platform. One of its games assesses the job candidates' speed, observational skills, effective or ineffective reactions to different patterns, and overall decision-making.

"The early signals are pretty interesting," says Marder. "There are not a ton of studies and use cases, but the early results are really promising. They allow companies to look at a more diverse talent pool and consider nontraditional talent. Now, it's like a blind audition."

Paskey says the HR industry is just getting started in measuring how effective machine-based assessments will be for hiring higher-skill-level employees. "I think that this is an age of disruption, in which many people and entities are working on this in the back room," he says. "Somebody's going to figure out how to use this data in a hiring model."

Will there ever be a day when human-free hiring is the norm? Paskey sees a future where some companies are more comfortable relying on human-free, machine-based hiring processes to hire, say, seasonal workers. "It may grow from there," he says.

Meanwhile, Marder says there's "really a great use for assessments as part of selection, and I think hopefully they are a supplement to human instinct, not a replacement for [it]. Because, depending on the role, sometimes you do need to sit across from a person and say, 'Can I work with them? Can I have them as part of my team?' "

"At the end of the day," Erickson says, "people are human, which means that, no matter how somebody does on their pre-hire assessment, if the hiring manager doesn't connect with them on some level, there's a really good chance that that person is not going to stay. [That's when you] start to do personality assessments on the hiring manager as well as the people who are being hired. I don't see computers taking over for this anytime soon."

Send questions or comments about this story to


Copyright 2017© LRP Publications