What Science Says About Hiring the Right Person

Predictive methods for hiring, for small and medium companies. 

As a psychology student, much of my day is defined by statistics. We do A LOT of statistical analysis, and it’s now the only way to convince me of anything. Some things are difficult to measure you might say… or maybe I should be thinking more about subjective information as well. In the case of business, however, and hiring in particular- I’m going to have to disagree. 

Sometimes we just get a feel for people’s personality, and maybe that should be enough to tell us whether someone is the right candidate for the job or not. WRONG!! You might fall into the trap of unconscious biases if you do this, and at worst you may make a terrible hiring decision, all because you believed that your traditional interviewing processes would work on their own.

Here’s how to keep hiring people you ‘click’ with, whilst also managing to make the right choices using the most scientifically proven predictive methods for hiring the right person. 

Let’s start with personality and cognitive testing… Personality Tests describe and predict the performance of applicants. The best model, called the Big Five Factor model, does this through measuring openness, conscientiousness, emotions, agreeableness, and neuroticism (Barrick et al, 1991). The results can then be used to aid you in your interview process, which is much more effective as a structured and personalised process. This is a win-win for you and the candidate, as it allows you the greatest prediction for performance, and the candidate the opportunity to show off the personality traits that mould their behaviour. 

Aside from the statistically strongest personality test, we also have Myers-Briggs. Though it is popular, it was never meant to be used for hiring, because it lacks valid psychometric properties (King & Mason, 2020). Psychometric properties are the key to results that can give you a clear picture of prediction, so the fact that they’re lacking from this model is troubling. Not really my cup of tea, and I wouldn’t recommend using it yourself for hiring purposes.

Analysing these predictive methods wouldn’t be complete without discussing cognitive ability tests too. They are very robust predictors of performance, and there’s an ample variety to choose from, sounds pretty good to me! You can see from the bar graph above that they are relatively good predictors for hiring, though the Big Five Model is still the reigning champion (Bertua et al, 2005). 

A short but important note regarding testing specific to recruitment consultancy roles- SPQ*GOLD is the perfect way to measure the drive, clarity, focus, resilience and comfort of candidates. This way, you’re quickly down to the nitty gritty, utilising a sales-specific behavioural assessment that translates ability on-paper into real time action for your company. It’s really a no-brainer if you want to prevent those colossal bad hiring decisions.

Think about any interviews you’ve had in the past when applying for a job. It was no surprise to me, at least, to learn that structured interviews, as already discussed, produce better predictions of job performance than traditional interviews do. But, did you also know that research shows that interviews carried out in person result in more accurate measures of job certainty that technology-mediated ones can (Blacksmith et al, 2016)? 

Yet more analysis of predictive methods for hiring the right person led me to an even more interesting discovery. The ineffectiveness of work experience as a scientific measure for hiring the right people shocked me at first. The reason being that work experience is one of the most commonly encountered measures of personal skills and is often assumed to be a good predictor of future ability in occupation-related positions. In fact, it has a certainty rating of 0.24 (Quińones et al, 1995, Schmidt, 2016)! That’s less than Myers-Briggs…

At school, I worked hard, listening to my teacher’s advice that good grades would help to secure me a good job in the future. At university, I strive for high results, convinced still that my education is central to my future occupational success. Statistics has told me once again that I am wrong. Though I, and many others I imagine, maintain the ideal that hard work at school will provide future opportunities for greatness, as a predictive method for hiring the right person, science shouts out in large volumes that education is a relatively poor method for predicting job performance (Schmidt, 2016). Maybe then, it is still a measure of our core personality traits that defines our determination to work hard, and our approach to learning is much more important than the grades we eventually flaunt on our CVs? 

Now that I have taken you on my journey of scientific investigation, here’s a nod to the future. Artificial intelligence is rapidly overwhelming many of our current disciplines, and eventually, it is bound to define psychometric testing too. But for me, for science, and for now, it seems necessary to hold out and make the most of our existing (and scientifically proven) personality tests, whilst we wait to see what gamification, now largely still in its’ adolescent form has to offer. Gamified assessments themselves collate the information used by existing psychometrics, then creating a new and engaging video game style test (Georgiou et al, 2019). Would your company be willing to risk the volatility of new and less researched materials, or are you more likely to stick with what science can tell us is the most reliable tool? 

Not everything can be explained by science, but it can always predict how efficient your next hire might be. Hopefully, these insights can help you to look at continued investment into your hiring processes because we are all aware of the real cost of a bad hire… The market for psychometrics is always changing, and you have all of the tools to keep pace. 

Author: Tallulah Goldsmith