> "That said, the results are based on self-reported answers"

Uhuh. I'm gonna go ahead and take this whole article with a massive chunk of salt; this breaks one of the main rules of how to do a statistically meaningful survey, so I don't see much value in the numbers.