Psychometrics and AI: How to Empower Talent Acquisition

“An invalid psychometric instrument is business poison, whereas a good psychometric is a litmus test”.
Share on facebook
Share on twitter
Share on linkedin

Get the best of
The Daily Bundle in your inbox every week

AI psychometrics

Get the best of The Daily Bundle in your inbox every week

Talent acquisition is difficult, which is why most companies have never been able to find the key formula that guarantees success in hiring. Looking for candidates, interviewing them, assessing their skills, and figuring out if they are the right fit for a company is a time-consuming process.

That’s one of the reasons why over 40% of businesses worldwide have outsourced most if not all of their hiring processes to specialized companies. Those who keep their recruitment in-house have to cope with either having a dedicated team of recruiters or risk overburdening their HR department with having to build a candidate pool.

What comes after is not a walk in the park either, choosing the right talent for your company requires patience and experience. To make matters worse, unscrupulous industry vendors offer packages, strategies, and “tricks” that go from the barely useful to the outright pseudo-scientific. 

Voice-recognition, lie detectors, body language assessment, dubious psychological testing, look hard enough and you’ll find someone offering tarot readings as a proven method of finding the perfect talent. 

Case in point, one of the most popular psychological assessment tools, the Myer-Briggs Type Indicator (MBTI for short), has been criticized time and time and again for its dubious psychometric properties and lackluster predictive value of job success. And still, it’s one of the most common personality tests used in hiring.

 

Psychometrics: The science of measuring the unseen

In the early 20th century, Charles Spearman, a British psychologist, noticed something quite unusual about student grades, those students who had higher grades in one subject tended to have higher grades in other subjects, even if those subjects had nothing in common. 

This went against the common notion at the time that a good grade was the natural result of learning and understanding the subject at hand. Using a statistical method that would later come to be known as factor analysis, Spearman found an “underlying construct” that explained what he had discovered, this construct is what we commonly refer to as “Intelligence” nowadays.

Long story short, Spearman’s extraordinary work lay down the foundations for psychometrics (the measurement of psychological variables) and differential psychology (the field of psychology specialized in finding differences between subjects and predicting the outcome of said differences). 

Historically, differential psychology and psychometrics have focused on intelligence and personality, but the field has expanded to include attitudes, motivations, values, self-concept, self-stem, and many other aspects of the psyche. 

The problem with psychological constructs, in general, is that we cannot see them directly, we can only infer their existence by measuring people’s behavior. There isn’t a single thing we can point to and say “that’s intelligence”, but we can point at someone’s outcome (like Spearman with students grades) and say “part of this outcome is due to that construct we call intelligence”.

By measuring the way people behave we can infer the value of the underlying construct, but just like any scientific endeavor, measuring these constructs requires calibrated tests, this process is called psychological validation. It can take months or even years before a psychological test is validated, and it’s a process that never stops. Thanks to cultural variation every time you export a test to another population you need to begin the process all over again. 

Is it worth the effort? the answer is yes, social intelligence and intelligence have shown to have strong predictive value with regards to academic and job success. In simple terms, people who do well on tests that measure these variables also tend to do well in their academic life and on their jobs.

Likewise, personality tests based on the Big Five model have shown that certain personality traits predict success in areas such as sales. Careful though, even if I’m a big proponent of psychometrics, I’ll be the first to admit that it’s not the end all be all of the talent acquisition. The evidence is there but it’s far from conclusive.

One of the big advantages of using personality and intelligence in tandem with other forms of assessment is that you can get a clear picture of someone’s analytical thinking, flexibility, and creativity before the interview phase. Sternberg’s triarchic ability test is a perfect example of a well-researched test that has shown a lot of promise.

 

AI and psychometrics: a match made in heaven

With the surge of available candidates and the globalization of the marketplace businesses are facing a completely different problem than a couple of decades ago. Before it was finding the right person for the job, now there is an overabundance of candidates leaving their digital profiles all over the internet.

Enter services like BairesDev’s Staffing Hero, an AI-powered solution that uses an algorithm to predict which skill set fits the client’s needs and builds the best engineering team from a talent pool of over a million profiles of technology experts. Clients just have to establish which variables they want, and the service handles the rest. 

AI-powered seemed like a thing from a dystopian novel up till a few years ago, and yet, here we are, and while there is undoubtedly room for improvement, we can’t also deny that AI is simplifying the process in ways that we would have never imagined. 

So, how does psychometry enter the picture? Interestingly, while fact-based psychology has always relied heavily on statistics, psychology as a whole has kept its distance from AI (which is ironic, since some of the most popular models of cognition are based on computer science).

But that is changing, psychometry as a field is on the verge of a revolution thanks to digital acceleration. With broader access to subjects to validate tests and with the integration with big data, psychologists are finding a solution to the age-old problem of finding a sample of sufficient size to make their models work. 

Predictive AI on the other hand has a lot to win from incorporating psychometric measures to the variables used to train the model. Differential psychology was founded in part out of the necessity to assess if candidates had the required potential to fill certain roles within the U.S army. And from that, it has grown into a respectable field with hundreds of tests designed to measure all kinds of human behavior.

Unfortunately, the marriage between the two is still a long way off, while many companies have used psychometric tools to assess candidates in the past, few, if any, keep copies of the data needed to train AIs, and without data, there is a lot of groundwork to lay down before we can start feeding psychological variables into our algorithms.

 

And yet there is hope…

More and more psychologists are seeking careers as data scientists, and software developers, in turn, have seen the benefits of having psychologists consultants on their teams, I honestly believe that the future of psychometrics is in AI and gamification and that software developers have a lot to learn from the rigorous methodologies of psychology and psychometrics.

As Dr. Knight Craig puts it “An invalid psychometric instrument is business poison, whereas a good psychometric is a litmus test”.

How useful was this post?

Click on a star to rate it!

Please enter a valid E-mail address.

Get in Touch

Access the Top 1% IT Talent, leverage our expertise to
jump-start your business.​

  • This field is for validation purposes and should be left unchanged.
If you previously need to sign a non-disclosure agreement, please email us at [email protected].
Scroll to Top

Get in Touch

Jump-start your Business with the
Top 1% of IT Talent.

Need us to sign a non-disclosure agreement first? Please email us at [email protected].

ACCELERATE YOUR DIGITAL TRANSFORMATION

By continuing to use this site, you agree to our cookie policy.