Introverts Need Not Apply
Everything Recruiters Get Wrong About Personality Tests
In theory, personality tests are like streetlights. They won’t make the path any shorter or easier, but they will help you see the way.
But when recruiters use personality tests bad things tend to happen. After almost a decade of recruiting, I am no stranger to mistakes. But, with each mistake, you learn more about what NOT to do. Personality tests are so badly misunderstood by the majority of recruiters that it’s no wonder that many companies are facing an increasing number of lawsuits for their misuse.
Here are some of the most common mistakes recruiters make when trying to use personality tests that in the majority of cases they don’t fully understand.
1- Using tests as the sole criterion for hiring or promotion decisions
When Deep Blue first beat chess master Garry Kasparov in 1997, it was heralded as the beginning of an age where computers were able to outsmart humans. In the 20 years since, however, the world of chess has come to a more nuanced understanding of how to use technology to improve results: harmoniously.
Today, the best chess teams in the world are created of a hybrid of humans and machines working together to produce the best results. Human-Machine hybrid teams consistently outperform teams of two humans or two machines.
This should be exactly the same with hiring. Relying solely on intuition is as foolhardy as relying exclusively on data gathered from personality tests. It’s the way the data can be used to inform intuition that makes a recruiter great.
2- Failing to set objective targets
Using Personality Tests to make your hiring process fairer and more consistent is a good thing to do. Using Personality Tests without first setting an objective measure is a pointless thing to do. First, you need to create a target goal, that is the same for every candidate applying for the role or company. Importantly you’ll need to be able to demonstrate that it specifically relates to the job at hand otherwise you run the risk of discriminating on the grounds of personality.
The most common mistake when it comes to not setting objective targets is accidentally implementing retrospective targets.
Without a pre-set narrative of what you are looking for in a candidate, many recruiters look at the results of psychometric tests and use them to confirm what they already thought about a candidate, while ignoring any new information given as inconsequential.
This is called confirmation bias, you can read more about it in hiring here.
3- Being too binary in your view of the data — e.g. “Introverts need not apply”
In 1948, psychology professor Bertram Forer gave his class a personality test, and without looking at the results, gave the entire class an identical feedback form, regardless of their answers. The average evaluation from the students was that the test’s accuracy was 4.26 on a scale of 5.
The reason that the results managed to appear to be so accurate was because they included phrases like: “Sometimes I am introverted, and sometimes I am extroverted”, which almost everyone could relate to.
One of the biggest problems with Myers-Briggs, and other personality type assessments is that a crude interpretation leads to people being put in boxes: you are either an “E” or an “I”, “P” or “J” etc. The data shows this simply isn’t the case.
When plotting aggregate results on a graph, you should see that people are bunched together, in groups of introverts and groups of extroverts (see bimodal distribution), showing there is a large difference between the “average introvert” and “average extrovert”.
However, when analysts looked at the data, they showed that it more closely resembles a normally distributed graph. This means that the line between introverts and extroverts is much more finely tuned, and differences are much smaller.
That’s cool, but how does it relate to recruiters?
By placing too much emphasis on whether a candidate is one personality type or another, you are likely to miss out on the nuances of their preferences and human behaviours. The difference between someone who is categorised as an “introvert” or an “extrovert” may only be one question in a 100 question survey, but the advice given to managers or recruiters may be completely different based on the label they were given.
4- Test manipulation AKA Social Desirability Bias
A willingness to impress means that it is often incredibly difficult to fairly assess candidates. This is genuinely a question I came across while researching this topic:
“How much do you agree with the following statement: “I am more competent than most people”- Not at all, not really, somewhat or definitely?”
While you definitely get what you pay for with online tests, the example above is still a valid hyperbole. Candidates know that every test they do in the recruitment process will be seen by their potential employers, and will do whatever they can to give what they perceive to be the right answer.
Tests for corporate values are particularly susceptible to this, often leaving companies open for candidates to give a false sense of themselves in order to impress. Just because a candidate answers a survey to say they are honest, respectful and hardworking, doesn’t mean this is factually true. Some tests mitigate for social desirability with control questions but generally, if your tests have answers which are either good or bad then alarm bells should be ringing.
The future of talent assessment
Personality tests aren’t perfect and it’s a wonder why we bother in the first place. Yet companies spend £billions each year on them.
We think there’s a better way.
Now forget the field of psychology ever existed for a second…
Seriously, try and forget it.
You have a problem that the people you’ve hired don’t fit in with how your people like to work. How would you solve that? You would try and measure how your team like to work in the first place.