Does the Gut or AI Make Better Personnel Decisions?

Artificial intelligence is making its way into personnel selection. What does this mean for team leaders and applicants?

Yannick is the leader of a team of software developers. They are missing a person for test engineering. Together with the HR department, he phrases a job advertisement.

The professional qualifications are quickly defined. Yannick finds it difficult to determine the required personal competencies. He decides on qualities that he values in himself and the other team members.

You Choose what You Know

At the end of the application process, there are two candidates to choose from. Both have excellent professional qualifications. The decision is not easy for Yannick, but he is convinced that he has made the right choice with Irina. "Anyone who ticks like me can't be wrong for this team," he thinks to himself.

The collaboration gets off to a great start, Irina does a good job. The only thing that bothers Yannick is that she chats and jokes a lot with the other team members. But he keeps Irina after the trial period.

Rude Awakening

He'd better not have, Yannick thinks to himself a few months later. Irina doesn't take diligence too seriously, increasingly questions his decisions and is not satisfied with her tasks. She wants to decide and shape things herself. Unfortunately, this is not what is planned for this position.

Friction increases, productivity decreases, and at some point, things start to go seriously wrong. Then the search for a new team member must start again after a short time.

Gut Decision Turns out to Be Wrong

What went wrong? Yannick was overwhelmed with the decision and made a gut decision for Irina – because she is similar to him. Now he has to realize that a quiet worker would have been better for the job.

What happened to Yannick happens frequently. Such miscasts cost companies a lot of money.

Known Hard, Unknown Soft Factors

How can Yannick do better next time? Decide with his mind instead of his gut? With some thought, he could have realized that there would be friction with Irina.

But people are complex and Yannick would have had to take into account countless other characteristics besides the ones mentioned above in order to make a rational decision.

However, he does not even know most of them. Only a few were evident from the application documents or came up during the interviews.

However, the personal and social skills are at least as important for the job as the professional ones.

How Do I Find out about the Soft Factors?

The hard factors, such as diplomas or professional experience, can be assessed well with the mind. The amount of data is manageable, and the information leaves little room for interpretation.

How does Yannick manage with the soft factors? We remember he specified some competencies for the job advertisement. Because he didn't quite know what to choose, he based it on himself and the other team members.

But more importantly, the competencies fit the job. If they had already been listed in the job advertisement, Yannick would have gotten a better picture of the candidates from the application documents. By means of a structured interview, he could have rounded this out in the job interviews.

Especially when filling key positions, an assessment center is also used to thoroughly determine the personal competencies of candidates.

The Problem of the Large Amount of Data

Having much more information at hand, would Yannick have been able to make a better decision with his mind than with his gut? Unfortunately, hardly. We humans find it very difficult to comprehend and analyze large amounts of data.

If Yannick had had to do this for a large number of applicants, he would have been overwhelmed with this task. Ultimately, he would have had to consult his gut again...

Can artificial intelligence help him?

First Steps

Using artificial intelligence (AI) is indeed a captivating idea. It easily manages to handle a lot of data and recognize patterns and correlations.

AI is already being used for personnel selection. Amazon, for example, used algorithms to search for people whose characteristics and qualifications most closely resembled those of its most successful employees. However, this went wrong because the program eliminated women. Algorithms are only as good as they are programmed to be.

Other systems try to read the suitability of candidates from text samples or the facial expressions of application videos. Or from the success of video games.

AI Systems Reach Market Maturity

"In general, the question can be asked whether there is enough and qualitatively sufficient data about the workforce in companies on the basis of which such systems can be built," wrote the Frankfurter Allgemeine at the end of 2019 on the subject of AI for personnel selection.

In the meantime, we've made more progress. Systems that use algorithms to record, evaluate and classify soft factors have reached market maturity.

Who’s Afraid of the Algorithm?

So will the algorithm soon decide whether I get the job I applied for? Is that ethically justifiable? If certain conditions are met, then from the point of view of Employees Switzerland, the use of algorithms is justifiable. It can even be advantageous for job applicants.

The conditions are:

  1. The algorithm must be fit for purpose, error-free and reliable.
  2. The system must be neutral, neither discriminating nor evaluating, nor collecting data that does not serve the purpose.
  3. All data must be treated confidentially.
  4. The system must record and match the competencies of both candidates and team colleagues at the future place of work. The information about them should be symmetrical.
  5. The final decision on hiring is always made by a human being, never by an artificial intelligence.

If these conditions are met, it is not the algorithm that decides whether a job is accepted or rejected. Candidates can even be sure that the job will then be a perfect fit.

It Takes Mind, Gut and Algorithm

So, is it better for Yannick to trust an algorithm than his gut and his mind when it comes to the next job? Not just, all three are important.

In terms of the hard factors, he best continues to lean on his mind.

With the help of the data provided by the algorithm, he can get a precise picture of the personal competencies of the candidates and find out how well they fit the job and the team. He has to supplement this with the impression he has gained from the application documents and the interviews.

Now he will have a different gut feeling than before, and it will help him make the right decision.

Author

Hansjörg Schmid

Hansjörg Schmid