Algorithms Make Better Hiring Decisions Than Humans

Guess what else algorithms can do better than humans? Everything. I guess we all better get used to a future full of middle management AI overlords.

An online test evaluated roughly one third (94,000) of the employees, who were asked questions about their technical skills, personality, cognitive skills, and fit for the job. Based on the results, an algorithm rated each applicant green (for high potential), yellow (for medium potential), or red (for low potential). On average, the employees who were evaluated by a machine stayed in the job 15% longer than those who were hired without being rated by an algorithm.

Not surprising. Hiring by humans seems more to do with how first impressions in terms of personality go rather than how they'll fit in the company. (Not saying that social interactions aren't a crucial part of hiring, but they tend to be the main part over a candidates actual capabilities or desire to work for a company.)

Steve said:

Guess what else algorithms can do better than humans? Everything.[/I]

Click to expand...

Even authoring algorithms?

It probably could.

That said, the algorithm is only as good as it's parameters. So if the parameters, suck, tha algorithm sucks.

In a low-skilled population where the average tenure was 99 days to begin with. Not sure how well this works up the skill chain.

TwistedAegis said:

In a low-skilled population where the average tenure was 99 days to begin with. Not sure how well this works up the skill chain.

Click to expand...

The other thing to remember... How do you get a raise today?

Rhetorical question, you don't even if you work long term at a company like say Intel

Well that question wasn't entirely rhetorical...the answer is you quit your present job and go work somewhere else and play "upgrade-itis" with your career upsetting the fruit basket frequently in the hope of more $$$ signs.....as being a loyal long term company employee has, for decades now, not netted you much of anything beyond a pat on the back.

So when a computer interviewing employees nets employees that stay 15% longer is that a good thing? Does it mean the computer hires better more reliable employees who are a benefit to the job and workplace? Or could it also mean the computer picks office drones tired or cynical of job upgrade-itis?

Nytegard said:

Not surprising. Hiring by humans seems more to do with how first impressions in terms of personality go rather than how they'll fit in the company. (Not saying that social interactions aren't a crucial part of hiring, but they tend to be the main part over a candidates actual capabilities or desire to work for a company.)

Click to expand...

no kidding, i suck at interviews. there were a lot of people doing my job that were not capable to do it. most of them got fire or left. only one to go!!! out of 50. lol.

~3 month retention time per employee, lol. So we're talking Walmart greeters here.

It all depends on the job and the questions. If the job is a dead-end low-level hell-hole, then an algorithm that looks for people that are just happy to have any job and they have a winning program.

considering a lot of jobs are filled because someone knows someone there are many less qualified people hired over more qualified people.

Interesting. I just listened to a program on NPR a couple of weeks ago in which this was discussed on one of the science programs. They said many companies are going this route to select better potential candidates for positions. One company used your zip code to determine to make it the next round. It was found the company had more success trying to hire employees that were within 20 minutes from the office. They tended to be more productive due to better work/ life conditions.

Over the last 10 years or so, the workplace and work force has changed. We are seeing new type of hiring practices and credentials to get hired in the workplace. Some of them can seem and actually be on a discrimating level. However, this also is a merger of technology and the new economy...whatever that is today.

BTW, the algorythym is nothing new. When I tested into a large company I was given a "personality" quiz. It represented about 50% of your qualifying score as it turned out. As long as you could read and perform basic math, you had your other 50%. The test was on paper but the computer scored it.

Sorry for the spelling. It's late and tired.

I see way too many hires based on someones appearances over their skills. also as mentioned way too many hires from a friend in the company (which now also can grant them types of immunities). So you end up with a workforce of low skilled friends and try to be a successful company.

Darunion said:

I see way too many hires based on someones appearances over their skills. also as mentioned way too many hires from a friend in the company (which now also can grant them types of immunities). So you end up with a workforce of low skilled friends and try to be a successful company.

Click to expand...

I've seen the exact opposite.

Companies hell bent on hiring over-qualified over-skilled applicants from out of town, over reliable good-enough temps that do the job well. My last place of work, one of supervisors was temp after his predecessor quit...he knew everyone, as he'd been working in the department lots over the years. He knew the job. Knew what it entailed. Knew all our gear. Got along great with everyone from bottom-ring all the way to the Top. He had been with the company and department for years, and was the temp supervisor for 12+ calendar months.

Why am I telling this story?

Because a search was setup for the full time job. He wasn't even considered. Why? Because a know-nothing outside committee demanded a master's degree was required to be in the running in the job app. Which you DID NOT need at all. Now everyone knew this guy could do the job, as he'd done it for a year. Everyone knew this guy was qualified and liked doing the work, he liked the company from bottom to top and was a great fit. The Boss even told him that if he had had a master's degree of any sort he'd have been an instant lock for the full-time position. Instead he was thanked for his service and sent on his way and an overqualified out-of-state guy was brought in.

The guy who got hired was a good guy. Like all new guys, it took him 6 months or so to get comfortable. 18 months after he got hired, he quit as he hated the micromanagement from the Home Office. And yet again another Search had to take place. They could have had an experienced guy who wanted to hang around who knew the people and the lay of the land....instead the Search went for over-qualification, and got a revolving-door position. That job locally is a bit of an inside joke actually now.

Funny thing, an algorithm is a human decision.. consistently applied.
Multi-factor & adaptive are too, just more complex I guess.

What is surprising about this?, nothing., its obvious the answer of what possibly makes a candidate stay longer was known, or well-guessed... the 'algorithm' just proved it, as it, just like a computer applied the same rules over and over over, unlike those pesky humans, with "feelings" and stuff like that.