I'd like to know how large their sample size was. I mean, this was yougov, so I expect at least some level of credibility to this, but depending on how large the same size is and how biased your selection is, you can get some really weird numbers.

E.g. do the same study with some old KKK members or with a school class in a black, impoverished neighbourhood or with a group of CEOs and you will get very different results.