The “1 in 5” probably makes it sound way more prevalent than it actually is.
Say you have 5 companies that interviewed 200 people each in the recent past
1 candidate had a parent come to their interview (which could mean “driving them to the interview and waiting in the lobby,” which is still weird but nowhere near the connotation of “sat in and listened to interview questions”)
1 in 5 companies will report they’ve had a parent come to an interview, even though 0.1% of candidates brought a parent
It’s impossible to create a survey that transmits people’s actual thoughts directly to you. Every single survey, ever, has flaws and biases. The game is figuring out the bias, how significant it is, and if you can get anything useful out of the results.
Thank goodness we have people willing to organically repeat what companies say with zero analysis, we don’t get enough of that in every major media outlet