Personality assessments can be a very valuable hiring tool.
But they’re not perfect in every way.
The main criticism is that nothing can really prevent job applicants from “faking” their responses.
Personality assessments never provide perfect predictions of job performance or other key outcomes—and perhaps this is due to applicant “faking.”
What if there was another way?
Perhaps we could use work references (such as current or former supervisors) to assess an applicant’s personality instead.
That seems like it could work. What does the research evidence show?
Personality psychology has a long history of using “informant ratings” to assess personality.
The “informants” might be parents, friends, school teachers, or even trained clinical psychologists.
All of these different sources can provide reliable and valid information, but there’s no reason to believe that any one source is necessarily more accurate than others.
And in general, people really do tend to agree with each other. For the most part, people see themselves quite similarly to how others see them.
There are some interesting ways in which people can disagree.
A recent large-scale analysis of research data discovered that student personality is much more important for academic success than was previously thought.
In fact, the analysis found that personality is even more important than a student’s IQ score. 4 times as important.
How did we miss this before?
It turns out that asking students to rate their own personality wasn’t the best strategy after all.
Much stronger predictions came from other people—asking teachers to rate the student’s personality, or even asking classmates to rate one another.
What about job applications?
There is much less research on informant ratings for job applications.
It’s an interesting idea, and potentially a very valuable one.
However, the research so far isn’t very promising.
A recent series of studies asked current or former supervisors to assess the personality traits of their subordinates—but under a few different conditions.
When supervisors were asked to rate one of their subordinates as part of a job application, they described the subordinate as significantly more motivated, self-disciplined, careful, and assertive.
In fact, many supervisors rated their subordinates more favorably than the subordinates rated themselves!
It seems that informant supervisors can be just as likely to try and “cheat” a personality assessment for a job application.
(In fact, it’s possible for the applicant’s own ratings to be less biased.)
But here’s the most interesting result from this series of studies:
None of the different conditions influenced the predictions made.
When using personality to predict actual performance, the context didn’t matter, and it didn’t matter whose ratings were used.
The supervisor’s ratings led to the same basic predictions as the applicant’s ratings.
And trying to cheat the assessment didn’t really work either.
Is it worth the extra effort to collect personality ratings from current or former supervisors?
Maybe. But then again perhaps not.