- Education And Career
- Companies & Markets
- Gadgets & Technology
- After Hours
- Banking & Finance
- Energy & Infra
- Case Study
- Web Exclusive
- Property Review
- Digital India
- Work Life Balance
- Test category by sumit
AI – The New Secret Sauce in Psychometric Assessments
Despite the proliferation of a variety of selection methods, hiring errors are still more commonplace than we would like them to be. Nearly 90% of the times, the hiring failures are explained by behavioral factors and not by a lack of technical know-how
Photo Credit : Shutterstock
Many disruptive technologies have been first thought up by science fiction writers whose technology inventions were only limited by their imaginations. The submarine, the cell phone and more recently, self-driving cars – all have a genesis in popular literature. One of the most pervasive themes in sci-fi literature and movies has been the portrayal of robots with artificial intelligence or AI. Remember R2D2 in Star Wars, Terminator, and Jarvis in Iron Man?
But guess what, that fantasy is already becoming a reality. Whether you are asking Amazon’s Alexa to play a particular song or Google’s Home to see if potato powder can substitute cornstarch in a recipe– AI can be found solving many of our problems.
Despite the proliferation of a variety of selection methods, hiring errors are still more commonplace than we would like them to be. Nearly 90% of the times, the hiring failures are explained by behavioral factors and not by a lack of technical know-how.
Given the logical structure and ease of deployment, Psychometrics remains the technique of choice for psycho-behavioral measurement in organizations. If you’re unfamiliar with psychometric assessments, then the best way to describe them would be any activity that is conducted in order to evaluate skills, knowledge, intelligence, personality or even educational achievement. Yes, even a school test is technically a ‘psychometric test’.
Perhaps the most well known psychometric test for workplace settings is the Myers Briggs (MBTI), which requires you to respond to a set of statements before you’re classified as one of the possible 16 personality types.Though, in more informed circles, the ‘Big Five’ model (the basis for tests like Psyft Personality Assessments) is thought to have superior explanatory power over the MBTI.
Faking– Is it possible in Psychometric Tests?
One of the biggest rubs with traditional psychometric tests had been that it’s easy to ‘look good’ on them. However, newer ipsative techniques, such as those used by Psyftand SHL, address the element of faking to a significant degree. However, the next round of disruption and innovation is expected to come with the advent of AI and is already in the works.
AI in Psychometrics – Use Cases
AI could potentially pull from a database containing a limitless number of data points. It could also take into account additional factors like delay in responding to a question, your mouse hovering over a particular answer before selecting another, or even cross-check against all your posts on social media. In fact, AI could even start to come up with questions dynamically to test in a way that no recruiter could ever dream of. Add some IQ assessments and throw in some technical questions, and we may soon wonder why humans were ever allowed anywhere near the hiring process.
Automating Entry Level Hiring
Hoping to be a pioneer in AI-based recruitment, Unilever has been using the technology to screen entry-level employees. Candidates play neuroscience-based games and have recorded interviews analyzed by AI. The results (for the US) – applications doubled, average time-to-hire went from four months to four weeks, recruiter's time spent on application review decreased by 75% and Unilever "hired their most diverse class to date." The company considers the experiment a big success and will continue it indefinitely.
To apply, candidates submit their LinkedIn profiles - no resume required. They then play 12 neuroscience-based games for about 20 minutes. The games are designed to assess traits like memory, risk preference/ aversion, focus and tendency to read emotional versus contextual cues.
If they qualify, they move on to a video interview, where they record responses to preset interview questions. Responses are then analyzed for things like keywords, intonation, and body language etc. All of this can happen on a smartphone or tablet.
Another interesting company with a focus on entry-level recruitment, is Headstart, a platform that uses algorithmic analysis to filter applicants. Its applicant matching system creates a detailed ‘fingerprint’ for every applicant using neural networks and machine learning, factoring in personality, interests, skills and demographic background in addition to traditional criteria such as qualifications and experience.
Assessments of the Future
An example of more futuristic research can be seen in Prof. Patrick Griffin's work where a collaborative problem-solving task is presented to at least two candidates, who must use separate computers. Each individual user receives an incomplete description of the problem. To solve the problem, users must work and collaborate using a messaging app.
As a part of the test, the platform records in detail all the onscreen activity, including the messaging bit, for both the candidates during the task, as opposed to simply recording their correct and wrong answers. All this data is then analyzed to gauge the abilities of the participants and allows observation of social components of collaborative problem solving, like perspective taking and task regulation.
AI is definitely the most exciting thing to happen to HR in a long time. But it is also important to remember that you don’t always believe what’s on the tin. Many of the current crops of AI solutions are riddled with inherent challenges.
For instance, the practical problems of massive data and processing power required by AI systems.To learn how to screen resumes as accurately as a human recruiter it needs several hundred to several thousand resumes for a specific role. And because AI is trained to find patterns in previous behavior, any human bias that may already be in your process – even if it’s unconscious – can be learned by AI.
The Last Word
So where does that leave the current set of psychometric tools? Well, these will remain relevant in the near future. Unless there appears a test that is 100% accurate, that doesn’t discriminate against anyone, that makes everyone feel good, that’s dirt cheap and doesn’t take any time at all to complete.
One thing is for sure though – assessment professionals would really need to up-tech their game. As they say, if you’re not at the table, you can be sure you’ll be on the menu.
Disclaimer: The views expressed in the article above are those of the authors' and do not necessarily represent or reflect the views of this publishing house. Unless otherwise noted, the author is writing in his/her personal capacity. They are not intended and should not be thought to represent official ideas, attitudes, or policies of any agency or institution.