There’s some interesting research being done at Cornell University to see if computers are better at detecting phony online reviews than people are.
The researchers asked a group of undergraduates and then a computer to determine whether 800 reviews of 20 Chicago hotels were legitimate. The results give us some insight into how the mind works and how it affects our judgment when it comes to interviewing job applicants.
In essence, the study underscores the influence of what is known as the “truth bias.” This is the tendency we have to believe that people are who and what they say they are.
We assume honesty unless shown otherwise
This is not only a common assumption, it’s a necessary assumption that operates in society – the default position that, unless we’re shown some reason to think otherwise, we generally believe we’re being told the truth. (Bernie Madoff comes to mind.)
As demonstrated by the undergraduates, we humans tend to assume reviews are honest unless shown otherwise.
On the other hand, once we find out the information is not true, we tend to assume many reviews are fake when they are not. They scored no better than chance in spotting the frauds.
On the other hand, the computer was able to identify the fakes about 90 percent of the time by looking at the words that fake reviewers used compared to the words actual reviewers used.
This same kind of thinking affects everyone involved in the interviewing process – interviewers and job applicants alike tend to believe what they are told.
As it evolves, this research may help us more effectively screen resumes and other claims made by potential employees. Right now, though, you can look for concrete terms and specifics versus generalized information when trying to discern if an applicant is being honest.
This was originally published on Mel Kleiman’s Humetrics blog.